Technology-delivered Physical Activity Program for Adolescents With Type 1 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Technology-delivered Physical Activity Program for Adolescents With Type 1 Diabetes (NCT05319600), a randomized na study. The primary outcome is Change From Baseline in Physical Activity.

2. Clinical Question

3. Trial Design

This was a randomized, two-arm, parallel-group trial with no masking. The registry lists the study as completed and records an enrollment of 31. The two arms were an intervention group receiving the Activate program and a treatment-as-usual control group. The intervention was delivered remotely across a 12-week period; control participants were instructed to wear a Garmin activity tracker and continue their usual daily diabetes care plan.

The registry lists the start date as 2022-04-18 and the primary completion date as 2023-07-20. The phase is listed as not applicable. The sponsor is the University of Vermont. These record details identify the study structure but do not explain how randomization was generated, whether allocation was concealed, or how study personnel handled assignment. Those methods are not reported in the registry record.

The registry's brief description frames the project as a feasibility, acceptability, and preliminary-efficacy trial. It describes the intended comparison as the program versus treatment as usual for behavior-change skills and daily active minutes, with additional interest in diabetes behavior regulation, psychological distress, inflammation, glycemic outcomes, and cardiovascular disease risk. The posted outcomes list one measure, change from baseline in physical activity, and one analysis. No secondary outcome results are posted in this record.

The outcome arm results report an intervention estimate with an associated count of 16 and a treatment-as-usual estimate with an associated count of 14. The record separately reports enrollment as 31; it does not reconcile enrollment with the arm-level outcome counts or explain withdrawals, missing tracker observations, or the population used for analysis. The analysis population and handling of incomplete observations are not reported in the registry record. Accordingly, the posted arm results should be read as the outcome information entered in the record, rather than as a complete flow of participants through the trial.

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:

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: 1 subject who withdrew from the study was excluded from analyses.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Physical Activity, measured over daily for all days from week 1 through week 12.

The primary endpoint is change from baseline in physical activity. The registry defines the measure as the average number of daily minutes with moderate-to-vigorous physical activity, captured by a Garmin fitness activity tracker. Its stated activity threshold is metabolic equivalents (METs) greater than or equal to 3 METs. The timeframe is daily for all days from week 1 through week 12.

Tracker-based measurement offers a way to quantify movement over many days rather than relying only on a participant's recollection. In general, repeated daily measurements can reflect day-to-day variation, while the average gives a summary of activity over the measurement period. The registry does not explain how days with missing tracker data were handled, how the baseline value was defined, or how the daily measurements were combined for analysis. Those details are not reported in the registry record.

The endpoint is relevant to the program's stated physical-activity aim, but it represents one specific behavioral measure. By itself, it does not establish whether the program changed other health outcomes, behavior-change skills, or feasibility and acceptability. Although the trial description mentions additional outcomes and mechanisms, no secondary outcome results are posted in this record. The recorded primary result therefore addresses the activity measure, not the full range of questions described for the project.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: it is measured in minutes, and its possible values are not limited to a set of categories. The registry describes the arm summaries as least-squares means and gives 95% confidence intervals. In general, a least-squares mean is a model-based estimate of a group mean. It may account for terms included in the fitted model, but the specific model terms used for these arm estimates are not reported in the registry record. The term should not be taken to imply adjustment for any particular baseline characteristic.

The reported comparison uses linear regression, with the result described as a slope. In general, a regression coefficient represents an estimated change on the outcome scale associated with a predictor or contrast in the model. For a two-group comparison, the coefficient can express the estimated difference between groups if the model is parameterized that way. The registry does not describe its coding, covariates, repeated-measure structure, or precise interpretation of the slope. The result of 129.2 should therefore be quoted as the reported regression slope rather than recast as an independently verified contrast.

A confidence interval communicates uncertainty in an estimate. Its width depends on the observed data and the assumptions of the statistical procedure. A narrow interval indicates greater precision than a wide interval, while an interval spanning values on either side of the null value permits a range of interpretations. For this outcome, the reported slope interval extends from -22.52 to 280.98. Because zero lies within that interval, the interval does not rule out a null comparison at the corresponding scale. That fact does not establish that the groups were identical; it means the estimate is imprecise enough to remain compatible with a range that includes no difference.

A p-value summarizes how unusual data at least as incompatible with a specified null hypothesis would be if the null hypothesis and model assumptions held. It is not the probability that the null hypothesis is true, and it does not measure the size or practical importance of an effect. The registry records p <0.05 for the slope. Because the listed 95% confidence interval contains zero, the p-value and interval do not align under the usual matched two-sided testing framework. The registry does not describe an alternative testing convention or explain this mismatch. The two summaries should be reported together, with the inconsistency made explicit rather than resolved by selecting one without explanation.

The outcome is based on daily activity-tracker observations across the stated period. Repeated measurements can contain variation across days within a participant as well as variation between participants. General statistical methods for repeated data may account for correlations among measurements from the same participant; treating correlated observations as independent can make uncertainty appear too small. The registry identifies a linear regression method but does not say whether it modeled day-level observations, participant-level summaries, or within-person correlation. The details are not reported in the registry record, so the analysis structure cannot be evaluated from the entry.

Missing tracker data can also matter for a repeated daily measure. In general, an analysis may use observed days, impute missing values, or apply another rule; conclusions can depend on the pattern and handling of missingness. The record does not report the number of missing activity days, the completeness threshold for a participant's data, or any imputation or sensitivity analysis. It is therefore not possible to assess how missing measurements influenced the estimates.

The registry describes the trial as randomized, which is a design feature intended in general to make treatment groups comparable at the start of a study. Random assignment does not ensure exact balance in a small sample, and it does not by itself address post-randomization missing data or deviations from assigned treatment. The record does not provide details on sequence generation, concealment, adherence, or crossover. These features cannot be evaluated from the posted information.

The study description presents the project as a feasibility, acceptability, and preliminary-efficacy trial. General power calculations are planning tools for designing future studies, not evidence about whether the current study had adequate power after the fact. No sample-size justification, target effect, or trial-specific power calculation is reported in the registry record. Any sample-size illustrations associated with the posted results should be understood as planning illustrations based on an observed estimate, not as observed power or proof that the trial could detect an effect of a particular size.

Finally, an outcome analysis is only one part of a trial's statistical interpretation. The record lists one outcome measure and one analysis, but does not provide a full analysis plan or identify the analysis population. It also does not report whether any interim analyses were performed. Those omissions prevent a full assessment of prespecification, multiplicity, and the relationship between the analysis reported and the original trial plan.

10. Analyses Reported by the Sponsor

The registry reports the sponsor's linear-regression analysis as a slope of 129.2, with a 95% confidence interval from -22.52 to 280.98 and p <0.05. It also gives least-squares mean arm estimates of 209.1 for the intervention group and 79.84 for treatment as usual, each with its own 95% confidence interval. These are the posted results and should be distinguished from a recalculated comparison.

No recalculated comparison is available for this continuous endpoint in the record. The registry notes that values were missing or not estimable for a two-arm comparison of a continuous endpoint, so there is no independent re-calculation to set alongside the sponsor's analysis. The sponsor's reported result therefore remains the only posted inferential comparison; it should not be described as confirmed by an independent calculation.

The sponsor-reported p-value and confidence interval warrant a specific caution. The interval includes zero, whereas the p-value is recorded as less than 0.05. Under the usual matched two-sided test and interval, those would ordinarily give consistent conclusions about the null value. The registry does not report why they differ. Neither statistic should be silently substituted for the other, and the record does not provide enough methodological detail to resolve the discrepancy.

The arm estimates and the regression slope answer related but not necessarily identical questions without details of model specification. The registry does not identify covariates, contrast coding, or the relationship between the arm estimates and the regression analysis. Interpretation should therefore retain the reported labels and uncertainty rather than infer unreported adjustments or methods.

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists one outcome measure and one analysis in the posted results, and no secondary outcome results are posted. This limits the visible results to the primary physical-activity measure, but it does not establish that the trial collected no other data or that no other analyses were conducted. The study description mentions additional outcomes and mechanisms, including behavior regulation, psychological distress, inflammation, glycemic outcomes, and cardiovascular disease risk. Results for those outcomes are not reported in the registry record.

Multiplicity arises when several outcomes, time points, subgroups, or analyses are examined. In general, conducting many tests can increase the chance of observing a nominally unusual result even when no corresponding effect exists. Researchers may address this through prespecified primary outcomes, adjusted testing, or clear labeling of exploratory analyses. The registry does not report a multiplicity strategy, nor does it provide a statistical analysis plan describing the hierarchy of outcomes or analyses. The single posted analysis should not be taken as evidence that there was only one analysis in the study.

Missing data are another important uncertainty. The primary measure relies on activity-tracker observations across days, but the record does not state how many observations were incomplete, how missing days were handled, or whether participants were excluded for insufficient tracker wear. It also does not report a missing-data assumption or sensitivity analysis. The analysis population is not specified. These omissions make it difficult to judge whether the estimates represent all randomized participants, participants with usable tracker data, or another group.

The record also does not report interim analyses or stopping rules, and it gives no detailed account of deviations from the planned analysis. Those items are not reported in the registry record; their absence from the entry should not be interpreted as evidence that a particular procedure was or was not used. Without a prespecified analysis plan and transparent missing-data reporting, the posted estimates are best treated as a limited account of the primary outcome rather than a complete statistical record of the trial.

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 reports zero serious adverse events in the intervention group, with 17 participants listed as at risk, and zero in treatment as usual, with 14 listed as at risk. It also reports zero other adverse events in each group, with the same group-specific at-risk counts. Deaths are reported as zero in both groups. The record contains no serious-event terms in its list of leading serious adverse-event terms.

These entries indicate that no events were recorded in the listed adverse-event categories for the groups and at-risk counts shown. They do not establish that the intervention has no safety risks. In general, an absence of reported events in a small study can coexist with uncertainty about uncommon events, and a registry entry may not convey the full detail of event collection, attribution, severity, or timing.

The record does not provide a narrative describing how adverse events were solicited or monitored, whether events were judged related to the intervention, or how the safety observation period was defined. Those details are not reported in the registry record. The adverse-event figures should therefore be read narrowly as the registry's posted counts, not as a comprehensive safety assessment or a comparison demonstrating equivalence between groups.

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

The result describes a comparison of change from baseline in daily moderate-to-vigorous physical activity between adolescents assigned to a remote skills-and-activity program and those assigned to treatment as usual. The reported arm estimates are higher for the intervention group, and the regression slope is positive. At the same time, the slope's 95% confidence interval includes zero and spans a wide range of values. The size and direction of the comparison are therefore uncertain based on the reported interval.

The p-value is recorded as p <0.05, but it does not align with the confidence interval under the usual matched two-sided inference framework. The registry does not explain this. The reported p-value alone should not be treated as a definitive resolution of the uncertainty, and the confidence interval should not be ignored. The record also lacks details needed to assess the model, missing tracker days, and analysis population.

This trial was described as addressing feasibility, acceptability, and preliminary efficacy. The posted primary outcome can inform the physical-activity question, but the registry does not post results for feasibility, acceptability, or the other outcomes described in the study summary. As a result, the posted record cannot answer the complete set of questions the trial was designed to explore.

The result does not, on its own, establish a causal effect with high precision, explain how any individual participant's activity changed, or demonstrate a change in glycemic or cardiovascular outcomes. Nor does the absence of adverse events in the registry entries establish absence of risk. These are limits on what can be concluded from the posted information, not a claim that the program did or did not produce any particular clinical benefit.

In general, a randomized comparison can help estimate differences between assigned groups, but confidence in the estimate depends on trial conduct, measurement, missing-data handling, and analysis choices. Several of those details are not reported in the registry record. The most defensible reading is therefore descriptive and cautious: the posted estimates favor a greater activity change in the intervention arm, but their uncertainty and the reporting limitations leave the magnitude unresolved.

14. Why This Trial Matters Statistically

This study brings together a behavioral program and objective activity tracking in a population identified in the registry as adolescents with type 1 diabetes. Statistically, it illustrates the value of considering effect estimates and their uncertainty together. A positive point estimate can appear encouraging in direction while a broad confidence interval leaves substantial uncertainty about the size and even the direction of the underlying comparison.

The result also illustrates why p-values should not be read in isolation. The registry reports p <0.05 while the corresponding 95% confidence interval includes zero. When summaries that are ordinarily linked do not align, transparent reporting of the model and test procedure is especially useful. The entry does not provide enough detail to resolve the difference, and the discrepancy itself is a reason to avoid overconfident interpretation.

The trial's stated feasibility and preliminary-efficacy goals matter for how its estimates are used. In general, early or feasibility-focused studies can help identify measurement and implementation questions and generate estimates for later planning. Their observed estimates can be imprecise, particularly when the record does not provide a detailed account of missing measurements or the analysis population. A planning illustration based on an observed estimate is not observed power and should not be confused with evidence that the original trial was adequately powered.

Finally, the record demonstrates the importance of complete reporting beyond the primary estimate. A clear accounting of participant flow, tracker-data completeness, analysis methods, secondary outcomes, and safety monitoring would make it easier to evaluate how robustly the posted result answers the trial question. The current record provides a useful starting point for statistical discussion while leaving these interpretation questions open.

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 arm descriptions are not equally detailed. The intervention description is truncated in the registry record, although the brief summary identifies Activate as a 12-week program combining behavior-change skills training and physical activity promotion. Consequently, the full content, intensity, and implementation of the program cannot be reconstructed from the posted details.

The registry reports an enrollment of 31 but gives outcome arm counts of 16 and 14. It does not explain the relationship between enrollment and those arm results, including whether participants withdrew, lacked usable tracker data, or were excluded from the analysis. No participant flow or analysis-population definition is reported. These are important gaps for interpreting a tracker-based endpoint.

The primary measure is described in terms of daily activity-tracker observations, but the registry does not report tracker wear requirements, data completeness, missing-day handling, or how daily data were reduced for analysis. It also does not specify baseline measurement details or the statistical model beyond linear regression for the comparison. Covariates, repeated-measure handling, and the meaning of the reported slope are not reported in the registry record.

The posted regression confidence interval includes zero while the recorded p-value is less than 0.05. The entry gives no explanation for this apparent inconsistency. The registry also does not provide a full statistical analysis plan, a multiplicity approach, or results for the other outcomes described in the study summary. These omissions limit assessment of prespecification and the wider study findings.

Finally, the registry identifies the study as randomized and unmasked but does not report the randomization sequence method, allocation concealment, adherence, or interim analysis procedures. The results should be considered in light of these reporting limits rather than as a complete account of trial conduct.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Technology-delivered Physical Activity Program for Adolescents With Type 1 Diabetes — NCT05319600.
• Registry record last updated 2025-03-26; status last verified 2025-03.
• Results first posted 2024-12-24.
• 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 describes a completed, randomized, unmasked, parallel-group trial of a 12-week technology-delivered physical activity and behavior-change program for adolescents with type 1 diabetes. Treatment as usual served as the comparator. The listed primary outcome is change from baseline in daily minutes of moderate-to-vigorous physical activity measured by a Garmin activity tracker.

The intervention arm estimate is 209.1, and the treatment-as-usual estimate is 79.84. The reported regression slope is 129.2, with a 95% confidence interval from -22.52 to 280.98 and p <0.05. The interval includes zero, making the apparent mismatch between the interval and p-value an important reporting caveat. No recalculated comparison is available for this continuous endpoint.

The registry reports zero serious adverse events, zero other adverse events, and zero deaths in both groups. It lists one outcome measure and one analysis, with no secondary results posted. Details on missing tracker data, the analysis population, the full statistical model, and a statistical analysis plan are not reported in the registry record.