Strategies to Reduce Antipsychotic-Associated Weight Gain in Youth: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Strategies to Reduce Antipsychotic-Associated Weight Gain in Youth (NCT00617240), a randomized phase 2 study. The primary outcome is Change From Baseline to Week 24 in Body Mass Index (BMI).

2. Clinical Question

The study is framed as a pilot test of whether introducing metformin along with second-generation antipsychotic treatment, accompanied by information about healthy eating and activity, could prevent or reduce weight gain and metabolic changes in adolescent youth. That purpose describes the research question; it does not establish a treatment effect. The registry identifies the main outcome as change in BMI from baseline to week 24, with change defined as the week-24 BMI minus baseline BMI.

The primary endpoint results report a mean BMI change of 1.62 kg/m^2 in the metformin arm and 1.8 kg/m^2 in the placebo arm. Both reported means are positive. These arm-specific summaries do not, on their own, establish whether the groups differed beyond random variation. The record does not provide an estimate of the between-group difference, its confidence interval, or a significance test. The clinical question is therefore only partly answered by the posted primary results: they describe the average change within each arm, while leaving the comparative effect unquantified.

3. Trial Design

This was a randomized, quadruple-masked, parallel-group phase 2 study. The registry reports enrollment of 9 and a completed study. Its start date is 2007-01, and its primary completion date is 2012-10. The sponsor was the University of North Carolina, Chapel Hill. The condition listed is weight gain.

The two registered intervention groups were metformin and placebo. The metformin arm description gives doses from 250mg to 2000mg/day for 26 weeks. The placebo is described as matched to metformin, with doses between 250/0mg and 2000/0mg per day. The primary outcome, however, is specified as change in BMI over 0-24 weeks, with the endpoint measured at week 24. The registry does not explain the difference between the 26-week arm description and the week-24 primary endpoint timeframe.

For the primary endpoint, the posted group summaries report 5 participants in the metformin arm and 4 in the placebo arm. The population definition for this endpoint is participants who took at least one dose of study treatment and had at least one post-baseline assessment. These endpoint group counts should be read alongside that definition; the record does not provide further detail about participant flow or reasons for any exclusions.

Randomization is intended to reduce systematic differences between assigned groups, while masking is intended to limit influences from knowing treatment assignment. Those general design features do not guarantee balance in a small study, nor do they replace a direct estimate of the treatment contrast. The registry record does not report the allocation ratio, the specific masking roles, or further details of the randomization procedure. It also does not report a sample-size justification or other design rationale. The study's pilot designation and recorded enrollment are relevant context when interpreting how much precision the endpoint estimates can offer.

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: The study sought to recruit participants who had minimal or no prior exposure to second generation antipsychotics and no previous treatment with metformin.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All participants who took at least one dose of study treatment and had at least one post 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 Change From Baseline to Week 24 in Body Mass Index (BMI), measured over 0-24 weeks.

The primary endpoint is change from baseline to week 24 in body mass index, measured in kg/m^2. The registry defines this as the BMI at 24 weeks minus BMI at baseline. BMI is described in the record as a measure of body fat based on height, weight, gender, and chronological age. A positive change therefore indicates that the measured BMI at week 24 was higher than at baseline; it is not a direct measurement of body fat.

BMI is used as a quantitative measure of change in body size during treatment. As a continuous endpoint, it preserves the magnitude of each participant's measured change rather than reducing the outcome to a yes/no category. The reported group means summarize the average changes for participants included in each arm's endpoint results. They do not describe every individual's trajectory, and they do not show how widely individual changes varied.

The primary endpoint population is defined as participants who took at least one dose of study treatment and had at least one post-baseline assessment. Thus, the results pertain to that recorded analysis population, not necessarily every person enrolled. The registry does not report how missing week-24 measurements were handled or whether another analysis population was used.

The reported means are 1.62 kg/m^2 for metformin and 1.8 kg/m^2 for placebo. The registry does not provide a two-arm contrast, confidence interval, or p-value for this endpoint. Consequently, the means describe observed arm-level changes but do not quantify the uncertainty around the difference between treatment strategies.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

For a continuous outcome such as BMI change, a mean is the arithmetic average of the measured changes among participants included in the group summary. It gives a compact description of the center of the observed values. A mean does not show whether most participants had changes close to the average or whether individual values varied widely. It can also be influenced by unusually large or small observations. The registry reports the primary endpoint as a mean, with change defined as week-24 BMI minus baseline BMI.

What does the standard error describe?

The standard error describes the precision of an estimated mean. In general, it reflects how much the estimate would vary across repeated samples under a statistical model; it is not the same as the standard deviation, which describes variability among individual observations. A smaller standard error indicates a more precise estimate of that mean, all else equal. Here, standard errors are reported for each arm's mean, but the record does not give confidence intervals for the means or for a treatment contrast.

The standard errors for separate arm means cannot be read as the standard error of the difference between those means. A comparative analysis needs to estimate the contrast directly and account for its uncertainty. The record marks the two-arm comparison for the continuous endpoint as unavailable because values were missing or not estimable. It therefore does not report the comparative estimate needed to assess the magnitude and precision of the difference. No contrast should be reconstructed from the arm-level summaries.

Why does the comparison matter?

In a randomized study, the main treatment question is generally comparative: how do outcomes differ between assigned groups? A within-arm mean change alone combines any treatment-related change with other influences, such as changes over time, concurrent care, measurement variability, or chance. A direct between-group contrast is designed to address the comparative question. Random assignment can support interpretation of such a contrast, but it does not make an unreported contrast available and does not eliminate uncertainty.

Confidence intervals are a general way to express uncertainty around an estimate. A confidence interval for the between-group contrast would show a range of values compatible with the data and the analysis assumptions, and would help readers assess both direction and precision. No such interval is reported in the registry record for the primary comparison. A p-value, if reported, would address compatibility with a specified null hypothesis under an analysis model, but it would not measure the size or clinical importance of an effect. No p-value is reported for this comparison.

What can planning calculations say?

Planning illustrations based on an observed effect are hypothetical design calculations, not observed trial outcomes. The stated assumptions for such illustrations are a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation; the target effect is described as a stated fraction of the observed effect. These assumptions define a calculation framework, not the study's documented original design. The registry record does not report that these assumptions were used to plan this trial, and no sample-size justification is reported.

Sample-size calculations illustrate how many participants per arm might be needed under specified assumptions to detect a target effect at a chosen power. They do not show the power actually achieved in a completed study. In particular, calculations using an observed effect after the trial cannot be treated as observed power or as proof that a study was adequately sized. The enrollment recorded for this pilot does not itself establish that the primary comparison could be estimated precisely.

The planning framework refers to an effect expressed as a fraction of the observed effect. Such a calculation depends on a chosen target and on assumptions about variability and analysis. The record does not provide enough information to treat any hypothetical planning result as the trial's own prespecified design. The calculations are useful only as illustrations of how planning assumptions affect required sample sizes; they do not alter the interpretation of the posted endpoint results.

What is not specified?

The registry record does not report the statistical model for the primary endpoint, adjustment for baseline BMI or other covariates, the method for constructing uncertainty intervals, or a hypothesis-testing procedure. It does not report whether interim analyses were conducted. These details should not be inferred from the fact that the study was randomized and masked. Nor should a particular analytic approach be assumed merely because the outcome is continuous.

As a general matter, the choice of analysis can affect an estimate's precision and how missing observations are handled. A repeated-measures approach, a single-timepoint comparison, and an analysis of change scores can rely on different assumptions. The registry identifies the endpoint as change from baseline, but does not describe a statistical model or make it possible to determine which modeling details were used. The safest reading is therefore limited to the posted arm-level means and standard errors, alongside the registry's explicit statement that the continuous two-arm comparison is not estimable.

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("Metformin", "Placebo")
n   <- c(5, 4)
m   <- c(1.62, 1.8)
sp  <- c(1.25, 0.5701)
sd  <- sp * sqrt(n)      # standard errors converted to SDs
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  Metformin  5  1.620  2.795  -1.851  5.091
CARM  Placebo  4  1.800  1.140  -0.014  3.614

11. Multiplicity, Missing Data and Analysis Populations

The registry lists four secondary outcomes: change from baseline to week 24 in insulin level, cholesterol level, and triglycerides, and incidence of metabolic syndrome. It also records 7 outcome measures. The primary outcome is the week-24 change in BMI. The record does not report a multiplicity adjustment, a hierarchy of hypotheses, or a prespecified strategy for interpreting multiple outcomes. Without such information, results across several outcomes could raise the general risk of chance findings when many comparisons are tested; whether and how that issue was addressed is not reported in the registry record.

The posted primary summaries do not include an estimable two-arm comparison for the continuous endpoint. The registry does not report comparative analyses for the listed secondary outcomes in the information. It also does not report an analysis plan explaining how primary and secondary outcomes were to be handled together. The number of registered outcomes should not be mistaken for the number of statistically tested hypotheses.

The primary analysis population is defined as participants who took at least one dose of study treatment and had at least one post-baseline assessment. The record does not report a full account of enrolled participants' progress through the study, the number lacking week-24 measurements, or reasons for missingness. It does not describe imputation, sensitivity analyses, or another method for handling missing outcome data. Consequently, the possible effect of missing data on the reported means cannot be evaluated from the registry record.

The record also does not report whether analyses followed intention-to-treat principles, used a per-protocol population, or used another defined population beyond the stated endpoint eligibility criteria. Those analysis-population details are not interchangeable, and the stated primary population should not be extended beyond its definition.

12. Safety Results

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

Other (non-serious) adverse events

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

Most frequent non-serious adverse event terms

The registry records 0 serious adverse events among 5 participants at risk in the metformin group, and 0 among 4 participants at risk in the placebo group. These are the serious-event figures recorded for the two arms. They describe the reported counts; they do not establish that serious events are impossible or rule out uncommon harms. The record does not list top serious-event terms.

For other adverse events, the registry reports that 5 participants in the metformin group were affected among 5 at risk, and 3 participants in the placebo group were affected among 4 at risk. These counts are distinct from the serious adverse event counts. The record does not provide event terms or further detail in the listed safety information, so the types, timing, duration, and severity of these other events cannot be described from it.

The death figures are not reported in the registry record. A missing death value should not be interpreted as a count of zero. Safety summaries also need to be read in light of the study's small recorded enrollment and the limited duration described for the intervention. Counts from a small study may offer little information about uncommon events, and these registry figures alone do not permit a broad assessment of safety.

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

The posted results establish what the registry reports for the primary endpoint: the mean BMI change was 1.62 kg/m^2 in the metformin arm and 1.8 kg/m^2 in the placebo arm, over a baseline-to-week-24 timeframe. Both mean changes are positive. The endpoint was defined for participants who took at least one dose of treatment and had at least one post-baseline assessment. The reported values describe that population's arm-level averages, not each participant's experience.

The results do not establish the size of the difference between arms. No comparative estimate, confidence interval, or p-value is reported, and the registry specifically indicates that values for a two-arm comparison of the continuous endpoint were missing or not estimable. The means therefore cannot support a conclusion about whether the observed groups differed statistically, how large a comparative effect might be, or how precisely such an effect is known. They also do not establish that metformin prevented or reduced weight gain.

The study's randomized, quadruple-masked, parallel design is relevant context, but design features do not replace an interpretable comparative result. The endpoint group counts and the recorded enrollment also underscore that the evidence is limited. A pilot study may be useful for describing feasibility or generating questions, but the registry does not report a sample-size justification or detail a broader analysis plan.

The registered purpose refers to metabolic changes as well as weight gain, and the record lists secondary metabolic outcomes. The primary results alone do not answer those secondary questions. The registry information does not provide comparative results for those outcomes or a multiplicity strategy.

These findings should be read as limited registry evidence about one primary measure at a specified timepoint. They are neither a definitive assessment of comparative effectiveness nor a complete account of benefits and harms. No treatment recommendation follows from the reported summaries.

14. Why This Trial Matters Statistically

This study illustrates a statistical distinction that is easy to miss: reporting a mean change in each randomized arm is not the same as reporting the randomized comparison. The means of 1.62 kg/m^2 and 1.8 kg/m^2 describe average within-arm changes in BMI. A treatment question, however, requires an estimate comparing those changes and a description of its uncertainty. The registry does not provide an estimable two-arm comparison for this continuous endpoint.

The distinction matters because readers can be tempted to compare two arm means informally. Without a direct contrast and its uncertainty, such a reading cannot show whether an apparent difference is precise, compatible with chance variation, or meaningfully different from no contrast under a specified model. The standard errors attached to the arm means do not supply the missing uncertainty for the between-group estimate.

The study also shows why outcome populations and timepoints need to be stated clearly. The primary population includes participants who took at least one dose and had at least one post-baseline assessment; the endpoint is defined from baseline to week 24, while the arm description gives a 26-week regimen. The registry does not explain this timing distinction or detail the handling of missing week-24 assessments. Those details constrain interpretation of the reported summaries.

Finally, the study's pilot designation, recorded enrollment, and absence of a reported sample-size justification are important statistical context. Planning illustrations based on observed effects can explore hypothetical designs, but they are not evidence of the original design's adequacy or of power achieved in the trial. The central lesson is to distinguish descriptive arm summaries, comparative inference, and hypothetical planning calculations rather than treating them as interchangeable.

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 primary comparative result is not estimable in the registry record, leaving the central between-group question unresolved. The arm-level means and standard errors do not supply a contrast, confidence interval, or p-value. As a result, the precision and statistical compatibility of a treatment difference cannot be evaluated from the posted primary summaries.

The endpoint population is restricted to participants who took at least one dose and had at least one post-baseline assessment. The registry does not report participant flow, the number with missing week-24 values, reasons for missing data, or the method used to handle missing outcomes. It also does not report whether the analysis population was defined in any other way. This limits assessment of how representative the endpoint summaries are of all enrolled participants.

There is a timing detail that remains unexplained: the metformin regimen is described as lasting 26 weeks, while the primary endpoint is change through week 24. The registry does not explain that distinction. It also does not report the primary statistical model, adjustment strategy, or planned handling of multiple outcomes. No sample-size justification is reported, and the pilot enrollment context limits the precision that can be inferred from the reported summaries.

The safety information is also limited. Serious adverse event counts are recorded, but deaths are not reported, and the listed information does not give terms or detailed descriptions for other adverse events. The registry does not provide comparative analyses for the secondary outcomes in the results. These limitations mean that the posted record cannot support a comprehensive appraisal of the study's metabolic outcomes, missing-data effects, or safety profile.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Strategies to Reduce Antipsychotic-Associated Weight Gain in Youth — NCT00617240.
• Registry record last updated 2014-03-11; status last verified 2014-02.
• Results first posted 2014-03-11.
• 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

NCT00617240 is a completed, randomized, quadruple-masked, parallel-group phase 2 pilot study titled “Strategies to Reduce Antipsychotic-Associated Weight Gain in Youth.” The official title is “Metformin Mitigation of Atypical Antipsychotic-Induced Metabolic Dysregulation in Adolescent Youth.” The sponsor is the University of North Carolina, Chapel Hill. The registry records enrollment of 9, a start date of 2007-01, and a primary completion date of 2012-10.

The intervention groups were metformin and matched placebo. The primary endpoint was change in BMI from baseline to week 24. The reported mean changes were 1.62 kg/m^2 for metformin and 1.8 kg/m^2 for placebo. The registry does not report an estimable two-arm comparison for this continuous endpoint. The primary population comprised participants who took at least one dose of study treatment and had at least one post-baseline assessment.

The registry lists secondary metabolic outcomes and safety counts. Comparative analyses for the secondary outcomes, missing-data methods, a multiplicity approach, and a sample-size justification are not reported in the registry record.