Exercise and Markers of Medial Temporal Health in Youth At-risk for Psychosis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Exercise and Markers of Medial Temporal Health in Youth At-risk for Psychosis (NCT02155699), a randomized na study. The primary outcome is Brain Volume.

2. Clinical Question

3. Trial Design

The registry record describes a completed study sponsored by the University of Colorado, Boulder. It lists the study as randomized and single-masked, while the model field says SINGLE GROUP. Those labels do not, on their own, explain the relationship between the dosage study and the randomized controlled trial components. The record lists phase as NA and enrollment as 50. The study start is recorded as 2014-07 and primary completion as 2022-07.

Three study arms are represented in the posted outcome summaries: Dosage Study Exercise, Dosage Study Exercise 2, and randomized controlled trial exercise and waitlist arms. The dosage descriptions specify exercise intensity and frequency over 3 months. The waitlist description specifies three months. The registry does not report the allocation procedure, the masking target, or how participants were assigned across the components. It also does not report a protocol explanation for the difference between the overall SINGLE GROUP label and the randomized allocation label.

The main endpoint is brain volume, with measurement described as change from pre-trial to post-trial. The record also lists Working Memory Assessment as a secondary outcome and reports 4 outcome measures. It lists 0 analyses. That does not establish that no analyses were conducted; it means analysis results are not posted in the registry record. The posted record contains descriptive summaries for the primary outcome but marks values as missing or not estimable for a two-arm comparison of this continuous endpoint.

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: Withdrew/excluded 3 Dosage (Exercise 2 group); 13 RCT (3 were excluded from the exercise group for not completing all 24 trials, 2 developed substance use disorders after baseline, 1 waitlist converted to a psychosis diagnosis during the protocol and was excluded); Final n=34; Change in hippocampal volume is reported below. On both measures negative scores indicate decreased hippocampal volumes over the study 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 Brain Volume, measured over pre-trial, post-trial (3-months).

The primary endpoint is titled Brain Volume and is described as change in mm cubed from baseline to follow-up, with the timeframe stated as pre-trial, post-trial (3-months). The target structures are the hippocampus and parahippocampal gyrus, both medial temporal structures. The endpoint description says the structures are delineated automatically using the FMRIB's Integrated Registration and Segmentation Tool within the FMRIB's Software Library image-processing suite. It also describes vertex analyses to assess post-trial shape changes across points on the structures.

Volume is a continuous measure: its values can vary across a range rather than being limited to categories. A continuous brain measure can describe the magnitude and direction of change in the measured structures. The registry also distinguishes the volume summary from vertex-based shape analyses, which are intended to show where a shape change occurs and what type of change is present. These measures address related but not identical features of brain structure.

The endpoint is relevant to the study's stated interest in medial temporal health. However, the registry record does not establish that a change in this measure would correspond to a particular clinical benefit. The posted primary outcome summaries do not include a reported comparative estimate or uncertainty interval. The record's wording gives the change unit as “baseline -followup”; the interpretation of positive and negative values should therefore follow that recorded convention rather than assuming an alternative sign direction.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What do the means and standard deviations describe?

For a continuous endpoint, the mean is a summary of the observed values in a group. The standard deviation describes how much individual values vary around that group's mean. The two summaries answer different questions: the mean describes the central location, while the standard deviation describes spread. Neither alone gives the precision of a comparison between groups. Precision for a treatment contrast is usually conveyed with an estimate and a confidence interval, and the registry record does not report those for the primary endpoint.

In this record, the outcome is change in brain volume in mm cubed, with the convention stated as baseline -followup. The reported arm means and standard deviations are descriptive. The small dosage-study groups, including the group with n=2, make those summaries particularly limited as a basis for generalization. Small groups may give unstable estimates of the mean and provide little information about the shape of the outcome distribution. The record does not include participant-level values or confidence intervals that could clarify this uncertainty.

Why does a two-arm estimate matter?

A randomized comparison aims to estimate how outcomes differ between assigned groups. The relevant quantity is not either group's mean in isolation but a contrast between groups, accompanied by an assessment of uncertainty. Randomization can support causal interpretation when the design, analysis population, and outcome handling are clear, but the registry's randomized label does not itself supply an effect estimate.

The record states that values are missing or not estimable for a two-arm comparison of the continuous endpoint. Thus, the posted summaries do not provide a comparative estimate, confidence interval, or p-value. The record also lists 0 analyses. These entries do not reveal whether analyses were conducted outside the registry, nor do they justify creating a new inferential result from group summaries. A recalculation from posted summary statistics, if one were possible, would be an unadjusted re-calculation and would not replace the sponsor's own analysis. The record does not report a sponsor comparison to take precedence.

What would a confidence interval communicate?

In general, a confidence interval gives a range of effect estimates compatible with the data and the statistical procedure used. A narrow interval indicates greater precision than a wide interval under the same method and assumptions. It is not a guarantee that the true effect lies in the interval, and it does not by itself establish clinical importance. Here, no comparative interval is reported, so the precision of an exercise-versus-waitlist effect cannot be assessed from the registry entry.

Why can over-dispersion matter?

In general, over-dispersion describes variability greater than a model expects. When a model underestimates variability, its standard errors can be too small and its intervals too narrow, making evidence appear more precise than it is. This is a general statistical concern, not a finding established by this record. The registry provides standard deviations for the arm summaries, but does not report a fitted model, distributional checks, or an analysis addressing over-dispersion. The observed summaries alone are not enough to diagnose a particular model problem.

How should planning illustrations be read?

In general, a sample-size calculation translates a chosen effect size and assumptions about variability, significance level, allocation, and analysis into a planning target. The registry's stated planning assumptions describe a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. Such illustrations are planning aids, not evidence about the study's achieved precision and not observed power. The registry record does not report the resulting planning values, the trial's own sample-size justification, or whether these assumptions were used in its design.

Sample-size targets based on an observed effect can be especially uncertain when the observed estimate is imprecise. Planning illustrations should therefore be interpreted as conditional on their assumptions rather than as a forecast of future findings. They do not repair missing comparative results in this record. The randomized component's posted summaries and the record's statement that a two-arm comparison is not estimable remain the relevant limits on interpretation.

What can be inferred from the outcome scale?

The primary measure is continuous and summarizes a structural feature. The endpoint description also includes vertex analyses, which assess shape changes across locations rather than only a whole-structure volume summary. Those methods may produce different kinds of results and should not be treated as the same endpoint. The registry does not post vertex-analysis estimates or analysis details. It also does not give a threshold for interpreting a volume change as clinically meaningful. Therefore, the posted means can be described statistically, but their clinical significance is not established by the registry record.

10. Confidence Intervals for Each Arm

No two-arm comparison is shown for this trial, but each arm's mean can be given with a t-based 95% confidence interval computed from the posted mean, spread and number of participants.

R code and output

arm <- c("Dosage Study Exercise", "Doseage Study Exercise 2", "RCT High Impact Interval Exercise", "RCT Wait List")
n   <- c(7, 2, 13, 12)
m   <- c(19.8, 54.8, -0.489, -6.28)
sp  <- c(50.8, 323, 0.104, 1.95)
sd  <- sp      # posted standard deviations
for (i in seq_along(n)) {
  se <- sd[i] / sqrt(n[i]); h <- qt(0.975, n[i] - 1) * se
  cat(sprintf("CARM|%s|%d|%.3f|%.3f|%.3f|%.3f\n", arm[i], n[i], m[i], sd[i], m[i] - h, m[i] + h))
}

CARM  Dosage Study Exercise  7  19.800  50.800  -27.182  66.782
CARM  Doseage Study Exercise 2  2  54.800  323.000  -2847.240  2956.840
CARM  RCT High Impact Interval Exercise  13  -0.489  0.104  -0.552  -0.426
CARM  RCT Wait List  12  -6.280  1.950  -7.519  -5.041

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 4 outcome measures and names Working Memory Assessment as a secondary outcome. It reports 1 secondary outcome posted. The record does not describe a prespecified hierarchy among outcomes, an adjustment for multiple comparisons, or a plan for handling multiplicity. When multiple outcomes are examined, the chance of a seemingly notable result arising by chance can increase; this is a general statistical issue. Without the analysis plan and outcome-level results, the impact of multiplicity cannot be assessed for this study.

The primary outcome description includes exclusions and reports a final n=34. It says that 3 participants were excluded from the exercise group for not completing all 24 trials, 2 developed substance use disorders after baseline, and 1 waitlist participant converted to a psychosis diagnosis during the protocol and was excluded. It also separately states “Withdrew/excluded 3 Dosage (Exercise 2 group); 13 RCT.” The registry record does not clarify the full disposition of participants across all study components or reconcile these details with the stated enrollment of 50.

The record does not report whether the primary analysis used an intention-to-treat population, a per-protocol population, or another analysis set. It does not report how missing outcome measurements were handled, whether any values were imputed, or whether sensitivity analyses were conducted. Excluding participants after randomization can affect the comparability created by random assignment; the magnitude or direction of any such effect cannot be determined from the posted information. The registry record also does not report a missing-data pattern or reasons for all missing measurements. These gaps limit assessment of both the analysis population and the robustness of the posted summaries.

12. 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 serious and other adverse-event counts separately for the dosage and randomized components. For the moderate-intensity dosage exercise group, 0 serious events and 0 other events are recorded among 7 at risk; deaths are 0. For the high-intensity dosage exercise group, the corresponding counts are 0 serious events and 0 other events among 2 at risk, with 0 deaths. In the randomized high-impact interval exercise group, 0 serious events and 0 other events are recorded among 13 at risk, with 0 deaths. For the waitlist group, 0 serious events and 0 other events are recorded among 12 at risk, with 0 deaths.

These entries describe events recorded in the registry; they do not establish that no unrecorded events occurred. Zero observed events in a group is not proof of zero risk, particularly when group sizes are limited. The registry lists no top serious-event terms. It does not report further detail about adverse-event definitions, collection procedures, severity categories beyond the serious/other distinction, or the period of safety monitoring. The counts should therefore be read as the posted safety record, not as a complete characterization of risk.

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

The registry record documents a study of exercise and medial temporal brain measures in youth at ultra-high risk for psychosis. It includes two dosage-study exercise groups and a randomized exercise and waitlist component. The primary endpoint is a continuous measure of brain-volume change, with the hippocampus and parahippocampal gyrus named as target structures. Descriptive means and standard deviations are posted for each listed arm.

Those summaries do not establish whether exercise changed brain volume relative to waitlist. The registry marks a two-arm comparison as missing or not estimable and lists 0 analyses. There is no posted comparative effect estimate or confidence interval from which to judge the size or precision of a between-arm difference. The available means should not be treated as proof of benefit, harm, or equivalence. Nor does the record establish that a measured change in brain volume would translate into a change in symptoms, cognition, functioning, or later clinical outcomes.

The study description expresses interest in symptoms, cognitive difficulties, and everyday social-occupational functioning, and the record names Working Memory Assessment as a secondary outcome. The registry does not provide enough analysis information to connect the primary structural measure to those outcomes. It also does not report a clinical-importance threshold for interpreting a volume change. The sign convention is recorded as baseline -followup, so claims about improvement or decline would require a defined interpretation of that convention.

The safety entries show no serious events, other events, or deaths recorded for the listed groups. They cannot establish the absence of risk, because event counts depend on the number of people observed, the monitoring period, and the completeness of reporting. In general, a registry record is one source of study information; the absence of posted comparative analyses and detailed methods leaves important questions unanswered. The appropriate conclusion is limited to what the record documents: descriptive outcome and safety entries, without an interpretable comparative efficacy result.

14. Why This Trial Matters Statistically

This study raises a statistical question about how to evaluate an intervention that is framed around a biological measure while also being motivated by clinical and functional outcomes. Brain volume is continuous and can be summarized by means and standard deviations, but a trial's central comparative question requires an effect estimate that respects the design and its analysis population. Descriptive group values are not a substitute for that contrast.

The record is also a useful example of why design labels need context. It records randomized allocation and single masking, while labeling the model SINGLE GROUP and reporting both dosage-study and randomized-component arms. Without an explanation of how those components were organized and analyzed, readers cannot confidently treat every listed arm as part of one common comparison. Clear reporting of design, allocation, analysis populations, and outcome definitions is necessary for an interpretable estimate.

Small group summaries and incomplete inferential reporting call attention to precision. Standard deviations communicate observed spread, but they do not provide uncertainty around a treatment contrast. A planning illustration can help translate assumptions into a future design target, but it cannot recover a missing estimate or establish what the completed trial demonstrated. The registry's lack of an estimable two-arm result makes this distinction especially important.

Finally, reporting exclusions and outcome counts matters because analysis populations affect interpretation. The record names exclusions and a final n=34, but does not give a complete account of missing-data handling or the analysis set. Transparent reporting of these elements would help readers understand the reliability and scope of any results. The statistical value of the record is therefore partly in the questions it makes visible: what comparison was intended, which participants contributed, and how uncertainty around the main outcome was assessed.

15. 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 registry record leaves uncertainty about the study structure. It identifies randomized allocation and single masking, but also labels the model SINGLE GROUP. It lists dosage-study arms as well as randomized exercise and waitlist arms, without explaining the relationship between the components or the assignment procedure. The arm description for the randomized exercise group identifies high-impact interval exercise, while the dosage descriptions give different intensity and frequency schedules. The record does not explain how the schedules correspond across components.

The primary endpoint summaries are descriptive, and the registry marks values as missing or not estimable for a two-arm comparison. No comparative effect estimate, confidence interval, or p-value is posted, and the analysis count is 0. The record does not clarify whether a comparison was planned but could not be estimated, or why the values are unavailable. It also does not report a protocol rationale, a sample-size justification, interim analyses, or a statistical analysis plan.

The outcome population information is incomplete. The primary description lists exclusions and final n=34, but the separate notation “Withdrew/excluded 3 Dosage (Exercise 2 group); 13 RCT” is not fully explained. The registry does not identify the primary analysis set or describe missing-data methods. The posted group summaries, including the dosage-study summary based on n=2, have limited capacity to characterize variability and precision.

Other limitations include the absence of reported confidence intervals and vertex-analysis results, no reported clinical-importance threshold for brain-volume change, and limited detail about safety monitoring. The registry also does not report how multiple outcomes were handled statistically. These limitations constrain interpretation of the posted information; they do not establish what additional analyses or documentation may exist outside the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Exercise and Markers of Medial Temporal Health in Youth At-risk for Psychosis — NCT02155699.
• Registry record last updated 2024-04-02; status last verified 2024-03.
• Results first posted 2024-03-13.
• 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.

19. Record Summary

The registry record describes a completed, randomized, single-masked study sponsored by the University of Colorado, Boulder, with enrollment listed as 50. It records the model as SINGLE GROUP and phase as NA. The study dates are 2014-07 for start and 2022-07 for primary completion.

The primary outcome is brain-volume change in the hippocampus and parahippocampal gyrus, measured from pre-trial to post-trial (3-months). Posted means and standard deviations are available for dosage-study and randomized-component arms, but a two-arm comparison is marked missing or not estimable. The registry lists 0 analyses, 4 outcome measures, and 1 secondary outcome posted, named Working Memory Assessment.

The primary outcome description reports a final n=34 and identifies exclusions, but the record does not fully clarify analysis populations or missing-data handling. No serious adverse events, other adverse events, or deaths are recorded in the listed groups. The registry does not post a comparative inferential result for the primary endpoint.