Healthy Hearts Healthy Minds: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Healthy Hearts Healthy Minds (NCT03373110), a randomized na study. The primary outcome is Change (Per Day) in Average Daily Steps From Baseline.

2. Clinical Question

The study rationale links depression and cardiovascular health with physical activity, and identifies motivation and energy as potential barriers to exercise. Its stated aim was to compare approaches designed to support physical activity, measured objectively with a Fitbit. That rationale explains the choice of a step-count endpoint, but it is not itself evidence that any intervention changed activity.

The primary question has more than one treatment group and more than one follow-up time. A complete interpretation would need to specify which group contrast and time point, or which pattern of change across time, represents the primary estimand. That specification is not reported in the registry record. The endpoint description alone cannot identify whether the intended focus was a comparison at 8 weeks, persistence at 16 weeks, or a joint assessment of both follow-ups.

3. Trial Design

Healthy Hearts Healthy Minds was a randomized, unmasked, parallel-group trial sponsored by Massachusetts General Hospital. The registry lists the study as completed, with enrollment of 506. The recorded start date is 2018-02-13 and the primary completion date is 2019-03-08. The phase is listed as NA. The registry summary says the project was funded by the Patient Centered Outcome Research Institute and involved the MoodNetwork and Health eHeart Alliance patient-powered research networks.

There were three behavioral study groups: online MBCT plus Fitbit, online CBT plus Fitbit, and Fitbit alone. The descriptions distinguish the online programs by their therapeutic approach. The MBCT description emphasizes awareness and meditation practice; the CBT description includes exercise goals, behavioral scheduling, rewards, and problem-solving around barriers. These are descriptions of the intended program content, not measurements of participant engagement or proof that a component affected activity.

Participants assigned to the therapy groups completed 8 weeks of online intervention sessions with biweekly questionnaires. The Fitbit-only group completed 8 weeks of biweekly questionnaires without therapy. After the first 8 weeks, no new online intervention material was provided to any group. The therapy groups continued to have access to material from the initial period, and all groups were instructed to continue wearing their Fitbit for the remaining 8 weeks. A follow-up assessment took place at week 16.

The registry calls the trial randomized but does not report the sequence-generation method, allocation concealment, or the ratio planned across the three groups. It also does not report the actual number randomized to each group in the design description. The safety record gives group-specific numbers at risk for adverse-event reporting, but those counts should not be treated as a full account of allocation or primary-outcome analysis populations.

Because masking is listed as none, participants and relevant study personnel were not described as masked. Step counts were recorded by a wearable device, but the registry does not report whether outcome processing or analysis was masked. The record also does not state the detailed statistical analysis plan or whether any protocol changes occurred.

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: 491 participants were consented, 361 participants met eligibility criteria and were randomized to study intervention arms. 21 participants withdrew their consent and, thus, data from these 21 participants are not used or presented. Of the 340 randomized participants, 314 participants had usable Fitbit data. Therefore, 314 participants were included in all analyses (i.e., primary and secondary outcomes). We have added adverse event data on all randomized participants (i.e., 361 participants).

Overall Study

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 (Per Day) in Average Daily Steps From Baseline, measured over Baseline, 8 weeks, and 16 weeks.

The primary endpoint is change from baseline in average daily steps, expressed as a change per day. Step counts offer an objective measure of movement recorded by the Fitbit, in contrast to a self-report of activity. The trial used this measure to assess whether activity changed during and after the online programs. The registry identifies baseline, 8 weeks, and 16 weeks as the relevant time points.

According to the endpoint description, participants in the therapy groups had 8 weeks of online sessions, while the Fitbit-only group had questionnaires without online therapy during that period. Afterward, participants in every group continued wearing the Fitbit for the remaining 8 weeks. This schedule makes the distinction between the end of the initial intervention period and the later follow-up important: an 8-week assessment concerns the period when new material was being delivered, whereas the 16-week assessment follows a period without new material.

The device records start and stop times for each day, including information such as when a participant was in bed. Steps were counted only during those recorded times. A day with no step-count data was treated as missing for that day. The record does not describe further processing rules, such as how days with incomplete wear time were handled, how daily observations were combined into an average, or how participant-level changes were calculated.

A mean is a measure of the center of a continuous outcome. A confidence interval expresses the precision of an estimated mean or contrast under the statistical method used; it is not a guarantee that an individual participant's outcome falls within that interval. Although the registry names a 95% confidence interval as the dispersion measure, numerical estimates and intervals for the primary endpoint are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The methods can be understood by separating the design, the outcome scale, the treatment contrasts, and the uncertainty around estimates. The registry describes a randomized, unmasked, parallel-group study with three groups and repeated measurement of average daily steps. It does not report enough analysis detail to reconstruct the statistical model or determine precisely which contrast was designated primary.

What does a mean represent?

Average daily steps are a continuous measure, and the registry identifies the primary parameter as a mean. A group mean describes the central level of the measured outcome for that group under a particular definition of time point and analysis population. A change-from-baseline mean concerns the difference between a follow-up measurement and baseline, summarized across participants according to the analysis method. Such a summary does not describe every participant's trajectory; individuals may have changes in different directions and of different sizes.

The phrase “change from baseline” also makes the baseline measurement part of the interpretation. It matters whether change is calculated for each participant and then averaged, or whether follow-up and baseline group means are compared. Those approaches can differ, especially when observations are missing or when baseline values vary. The registry record does not specify the computational approach, adjustment variables, or model used for the primary outcome.

Why do the time points matter?

The outcome schedule includes baseline, 8 weeks, and 16 weeks. The timing corresponds to an initial period with online intervention material for the therapy groups and a later period in which no new material was introduced. An estimate at 8 weeks addresses activity around the end of the initial program period. An estimate at 16 weeks addresses activity at follow-up after the groups had continued wearing their devices.

Repeated measurements on the same participant are generally correlated: a person's step counts at different time points are not independent observations. General statistical practice accounts for that dependence when estimating change across time. The registry record does not report whether a repeated-measures model, separate time-point analyses, or another approach was used. It also does not report whether the analysis tested an overall time-by-group pattern or evaluated particular group contrasts at particular visits.

How should confidence intervals be read?

A confidence interval communicates uncertainty around an estimated quantity. A narrower interval indicates greater precision under the assumptions of the method, while a wider interval indicates less precision. It is not a range that contains a fixed proportion of individual participant outcomes. The registry specifies a 95% confidence interval for the primary mean, but numerical confidence intervals are not reported in the registry record. Precision therefore cannot be assessed for the endpoint.

In a treatment comparison, the interval for a contrast is especially informative because it describes uncertainty in the difference between groups. A set of separate group means does not by itself show the uncertainty in their difference, and visual separation of estimates is not a substitute for a contrast with an interval. No such primary contrast is reported in the registry record.

What does randomization contribute?

Random assignment is intended to make treatment groups comparable on average with respect to measured and unmeasured baseline characteristics. In general, this supports a causal interpretation of a well-conducted comparison when follow-up, outcome measurement, and analysis preserve the randomized comparison. Randomization does not ensure identical groups in a particular trial, prevent missing data, or correct problems caused by departures from the assigned treatment strategy. The record does not report the randomization implementation details or the analysis population needed to evaluate these issues.

What can planning illustrations show?

Sample-size illustrations are planning exercises rather than observed trial findings. The registry describes an illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the target effect is expressed as a fraction of the observed effect. These assumptions describe the illustration, not necessarily the original trial design. The registry does not report the trial's own sample-size justification, and an illustration based on an observed effect does not represent observed power. The illustrations cannot supply the missing primary outcome estimates or determine whether the actual design was adequate.

For continuous outcomes, general sample-size calculations depend on the effect to detect, the variability of the measure, the chosen error rate, and the planned allocation and analysis. When repeated measurements or multiple treatment contrasts are involved, the model and multiplicity strategy may also matter. The registry record does not report a complete design calculation for the trial. No sample-size value from the illustrations should be read as a count of participants actually analyzed.

Why can step-count data be difficult to analyze?

Daily activity data may vary substantially from day to day, and the number of observed days can differ among participants. In general, over-dispersion means that observed variability exceeds what a simple model assumes; ignoring it can lead to confidence intervals that are too narrow and tests that are poorly calibrated. This is a general consideration for count-like measures, not a claim that over-dispersion was present in this trial. The registry identifies the endpoint as a mean of a continuous measure and does not report a distributional assessment or model diagnostics.

Device-based measurement also requires rules for valid wear and missing periods. Here, the record says that the device's recorded start and stop times determined when steps were counted and that a day without step data was treated as missing. It does not report the number of valid days required for a participant-level average, the treatment of partial days, or whether missingness was modeled or imputed. Those details affect how the reported outcome would be interpreted, but they are not available in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The primary endpoint has repeated time points and three study groups, while the registry also lists secondary outcomes stratified by PHQ-9, ASRM, WHO-5, PSS, age, sex, anxiety, and smoking. Multiple outcomes, time points, group contrasts, or subgroup analyses create a multiplicity issue: the more questions tested, the greater the chance of at least one apparently notable result arising by chance if each is assessed separately. General approaches include prespecifying a primary contrast, adjusting for multiple testing, or interpreting secondary analyses as exploratory. The registry record does not report a multiplicity strategy or identify one numerical primary comparison.

The record lists 14 outcome measures and 6 analyses. These counts do not identify which analyses were confirmatory, which contrasts were tested, or whether any adjustment was applied. The statistical analysis plan, prespecified hierarchy, and any interim analyses are not reported in the registry record. It would therefore be inappropriate to infer that the posted endpoint was tested with a particular correction or that secondary findings would have controlled error rates.

For missing data, the outcome description says that a day with no step-count data was treated as missing for that day. It does not report how much daily data were missing, whether missingness differed by group or time point, or whether missing measurements were assumed to be missing at random or handled by a particular model or imputation method. It also does not report how incomplete participant-level follow-up was treated. These omissions limit assessment of possible bias and precision.

The analysis population is not reported in the registry record. In particular, the record does not specify whether the primary analysis followed all randomized participants according to assigned group, included only participants meeting a minimum device-use threshold, or used another rule. Nor does it give analysis denominators at each visit or explain how participants with no usable baseline or follow-up data were handled. The enrollment count of 506 should not be treated as the number analyzed.

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: angina (0.7%).

The registry reports serious adverse events for the three groups, with the number at risk and the number affected. In the MBCT plus Fitbit group, 144 were at risk for serious events and 0 were affected. In the CBT plus Fitbit group, 145 were at risk and 1 was affected. In the Fitbit-alone group, 72 were at risk and 0 were affected. The record reports 0 deaths in each group.

For other adverse events, the registry lists 144 at risk and 53 affected in the MBCT plus Fitbit group, 145 at risk and 45 affected in the CBT plus Fitbit group, and 72 at risk and 30 affected in the Fitbit-alone group. These are counts of affected participants as presented in the safety record. They are not estimates of treatment effect, and they do not establish whether an event was caused by an intervention.

Angina is listed among the top serious terms at 0.7%. The registry information does not provide further detail in this record about its timing, clinical course, or relationship to study participation. Serious adverse-event counts are distinct from the broader “other” adverse-event counts. Differences in counts should not be interpreted as comparative safety evidence without the underlying definitions, ascertainment procedures, follow-up, and analysis.

Safety counts also have their own denominators and should not be substituted for the number included in the primary step-count analysis. The registry record does not report the detailed adverse-event collection methods or group-specific duration of safety follow-up.

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

The registry establishes that this was a randomized, three-group trial of online MBCT plus Fitbit, online CBT plus Fitbit, and Fitbit alone, with average daily steps as the primary outcome. It describes the timing and broad structure of the interventions and follow-up. It does not report the primary outcome estimates needed to quantify how much activity changed or whether the groups differed.

That distinction is important: an outcome being named in a registry is not the same as a result being available. Without group estimates, contrasts, confidence intervals, and analysis denominators, the size and precision of any observed effect cannot be judged. The missing numerical primary results do not demonstrate no difference, and they do not support a claim that one approach was superior.

The endpoint is an objective device-based measure, but objectivity does not remove all measurement challenges. Step recording depended on device-recorded start and stop times, and days without step data were considered missing. The registry does not explain how much data were missing or how incomplete records were handled. This leaves uncertainty about how representative the analyzed measurements would be of participants' daily activity.

The study's design also distinguishes the initial 8-week program period from follow-up at 16 weeks. These time points address different stages of participation. Because the registry does not specify a single primary contrast across treatment groups and time points, readers should avoid treating one time point as definitive or combining the time points without a reported analysis method.

The sample-size illustrations are planning calculations based on the observed effect and stated assumptions, not results that describe trial performance. They are not observed power. The registry record does not report the trial's own sample-size justification. Similarly, adverse-event counts provide a descriptive account, but do not establish that an event was caused by treatment or that one group was safer.

Accordingly, the record supports a description of the research question, design, measurement schedule, and reported safety counts. It does not support a conclusion about the primary intervention effect on average daily steps. Further interpretation would require the actual primary estimates, their uncertainty, the analysis population, missing-data methods, and the prespecified comparison.

13. Why This Trial Matters Statistically

This trial is statistically notable for pairing randomized treatment assignment with a wearable measure of daily activity. An objective measure can reduce reliance on participants' recall, while repeated observations can describe activity at the end of the initial program and at follow-up. Those strengths still depend on clear rules for valid measurement, missing data, and analysis.

The design contains three groups and repeated outcome measurements. This structure makes the choice of estimand consequential: the scientific question could concern one therapy compared with Fitbit alone, a comparison between the two therapies, a particular follow-up time, or changes over the full schedule. A transparent primary contrast helps readers distinguish the main test from additional analyses. The registry record does not report that contrast or a multiplicity plan.

Statistical precision matters as much as the direction of an estimate. A point estimate without a confidence interval does not show how uncertain the estimate is, and a nominally favorable result would not by itself address the number of comparisons. Here, the primary estimates and intervals are not reported in the registry record, making assessment of effect magnitude, precision, and robustness impossible.

The device-based endpoint also illustrates how missingness can shape an analysis. A missing day is not necessarily equivalent to a day with no activity. If step data are unavailable for reasons related to activity or intervention engagement, the observed days may not fully represent the intended outcome. Reporting the amount and pattern of missing data and the handling strategy is therefore important for evaluating a wearable-based endpoint.

Finally, enrollment and safety counts do not answer the primary efficacy question. The trial enrolled 506, but the number contributing usable measurements at each time point is not reported in the registry record. The statistical value of this study can be fully judged only when results, analysis populations, uncertainty, and analytic choices are reported together.

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 main limitation for interpreting the trial from the registry record is the absence of numerical results for the primary endpoint. No group-specific means, changes, treatment contrasts, confidence intervals, or p-values are reported. The direction, magnitude, and precision of the primary outcome therefore cannot be assessed.

The primary measure spans three time points and three groups, but the registry does not identify a single primary contrast. Without that information, the intended comparison and the role of the 8-week and 16-week assessments remain unclear. The record also does not describe the statistical model, adjustment variables, or handling of within-participant correlation across visits.

Step counts were limited to periods defined by device-recorded start and stop times, and days without step-count data were treated as missing. The record does not report wear-time requirements, the amount or distribution of missing data, or the approach to incomplete follow-up. These omissions make it difficult to judge the representativeness of observed step counts and the potential effect of missingness.

The registry lists multiple secondary outcomes and analyses but does not report a prespecified testing hierarchy or multiplicity adjustment. It also does not report the analysis population or participant flow by group and time point. Although enrollment was 506, this does not reveal how many participants contributed to the primary analysis.

Further design details are not reported in the registry record, including the randomization implementation, allocation concealment, detailed intervention adherence, and whether outcome processing or analysis was masked. The available safety counts are descriptive and do not include enough information to infer causality or make a comparative safety assessment. These limitations concern what can be concluded from the registry record; they do not establish that the unreported methods were absent.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Healthy Hearts Healthy Minds — NCT03373110.
• Registry record last updated 2022-02-02; status last verified 2022-01.
• Results first posted 2021-04-29.
• 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

Healthy Hearts Healthy Minds was a completed, randomized, unmasked, parallel-group study sponsored by Massachusetts General Hospital. The registry lists enrollment of 506 and the conditions cardiovascular diseases and depression. The recorded start date was 2018-02-13, and the primary completion date was 2019-03-08.

The groups received online MBCT plus Fitbit, online CBT plus Fitbit, or Fitbit alone. The primary outcome was change from baseline in average daily steps, with measurements at baseline, 8 weeks, and 16 weeks. The record describes 8 weeks of online intervention material for the therapy groups, followed by continued Fitbit use through week 16.

Numerical results for the primary outcome, the specific primary contrast, the analysis population, and the missing-data method are not reported in the registry record. Safety counts are reported by group, including one serious adverse event in the CBT plus Fitbit group and no deaths. The registry's sample-size material is described as a planning illustration based on the observed effect, not as the trial's original design justification or observed power.