Pioglitazone as a Treatment for Lipid and Glucose Abnormalities In Patients With Schizophrenia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Pioglitazone as a Treatment for Lipid and Glucose Abnormalities In Patients With Schizophrenia (NCT00231894), a randomized phase 4 study. The primary outcome is Change From Baseline in Serum Triglycerides at 3 Months ( 3 Months-baseline) US Sample.

2. Clinical Question

3. Trial Design

The registry record identifies this as a completed phase 4 randomized trial with a parallel-group model and quadruple masking. Enrollment was 56. The record lists a start in 2005-05 and primary completion in 2010-01. The sponsor was Nathan Kline Institute for Psychiatric Research.

The two described arms were pioglitazone plus lifestyle diet and placebo plus lifestyle diet. Pioglitazone was described at 30-45 mg daily, while the comparison arm used daily placebo capsules. The brief summary describes treatment for 12 weeks. It also describes assessment of fasting glucose and lipids, glucose-tolerance tests, a neurocognitive battery, and verbal memory at baseline and during treatment.

The primary outcome was change from baseline in serum triglycerides at 3 months in the U.S. sample. The endpoint description identifies participants with baseline and three-month triglyceride values and lists 25 in the pioglitazone arm and 19 in the placebo arm. These endpoint-specific counts should not be confused with overall enrollment. The record does not report the full analysis population definition or explain how participants without endpoint measurements were handled.

Conditions listed include diabetes, schizophrenia, insulin resistance, and cognitive impairment. The registry record does not report details of the masking procedures, the randomization method, or the original sample-size justification. It also does not report the trial's full statistical analysis plan or whether any interim analyses were planned.

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: Data is from participants at U.S. sites N=44, Pioglitazone =25, Placebo =19, who had values of triglycerides at baseline and three months followup. See table 2 of published paper given in reference for further details.

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 in Serum Triglycerides at 3 Months ( 3 Months-baseline) US Sample, measured over pre-treatment and during 3 months of treatment.

The primary endpoint is change from baseline in serum triglycerides at 3 months, measured in mg/dL, among participants at U.S. sites. Change scores describe how a participant's measured value differs from that person's baseline value over the stated follow-up. Comparing changes between randomized arms can describe whether the average change differed between the assigned treatment strategies over that period.

Triglycerides are one component of the lipid outcomes identified in the study's focus on lipid and glucose abnormalities. This measure is used as a continuous outcome, so the posted arm values represent least-squares means rather than event counts. The registry record labels the accompanying dispersion as a standard error. The endpoint description states that the U.S. sample includes participants with triglyceride measurements at baseline and three-month follow-up.

The primary outcome is specific to that population, measure, and follow-up. It does not by itself describe every participant enrolled, every lipid or glucose measure, or cognitive outcomes. The registry record does not provide a detailed rationale for selecting this particular endpoint as primary.

7. Results Posted in the Registry

• Pioglitazone: mean -49.65 versus 30.99 on control; difference -80.64 (95% CI -151.19 to -10.09), standardized effect d = -0.68; z-test p = 0.025.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Change in RBANS List Recognition Scores at 3 Months (3 Months-baseline) US Sample

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Pioglitazone: mean 0.27 versus -1.11 on control; difference 1.38 (95% CI 0.20 to 2.56), standardized effect d = 0.69; z-test p = 0.022.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Pioglitazone than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry lists secondary outcomes spanning cognitive performance, glucose, triglycerides, glucose tolerance, and HDL. A posted result is available for change in RBANS List Recognition scores in the U.S. sample. The other listed secondary outcomes do not have results detailed in the registry record. Consequently, the record does not support a broad comparison of patterns across all secondary measures.

The available cognitive outcome addresses a different domain from the primary triglyceride endpoint. It can provide complementary information about the study's wider aims, but it does not confirm or explain the primary lipid result. Likewise, the list of additional metabolic outcomes indicates that several dimensions were considered, but without their posted results it is not possible to determine whether they align with the primary comparison.

Secondary results generally carry less weight than a prespecified primary result when many outcomes are examined. Each additional comparison creates more opportunity for an apparently notable result to occur by chance. The registry record does not report whether the secondary outcomes were adjusted for multiplicity, whether a hierarchy was prespecified, or how missing values were handled. Any unadjusted secondary finding should therefore be treated as exploratory rather than as independent confirmation of the primary endpoint.

The available registry information also does not give a complete analysis plan or describe all endpoint-specific populations. That limits interpretation of whether differences across outcomes reflect real variation, analysis choices, or missing reporting. The secondary outcome list helps define the scope of the trial, but it does not substitute for reported estimates and uncertainty for each measure.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Change in RBANS List Recognition Scores at 3 Months (3 Months-baseline) US Sample
• Change From Baseline in Serum Glucose at 3 Months ( 3 Months -Baseline) China Sample
• Change From Baseline in Serum Triglycerides at 3 Months (3 Months-baseline) China Sample
• 2 Hour Glucose From Glucose Tolerance Test China Sample
• Change From Baseline in Serum HDL at 3 Months (3 Months-baseline) China Site

9. Statistical Interpretation

10. Statistical Methods Explained

What does a least-squares mean represent?

A least-squares mean is a model-based estimate of an average outcome. In general, such estimates may account for variables included in a statistical model. The registry record labels the endpoint values as least-squares means but does not report the model specification, covariates, or adjustment procedure. The arm estimates should therefore be read using the information available in the record, without assuming which factors were included in the model.

The posted primary arm values are -49.65 mg/dL for pioglitazone and 30.99 mg/dL for placebo. The comparison value of -80.64 mg/dL is a simple unadjusted recalculation from these posted summaries. It is expressed as the pioglitazone value minus the placebo value, so a negative value represents a lower average change in the pioglitazone arm.

What does the standard error describe?

A standard error describes the statistical uncertainty in an estimated mean. It is not the spread of individual participant measurements. Here, the record reports standard errors of 24.01 and 26.82 for the arm means. In general, an estimate with a smaller standard error is more precise, while a larger standard error indicates more sampling uncertainty. The standard error alone does not show how much individual triglyceride values varied.

How should the confidence interval and p-value be read?

A confidence interval describes a range of estimates compatible with the data and the method used, subject to its assumptions. The primary recalculated contrast has a 95% confidence interval from -151.2 to -10.09 mg/dL. A p-value describes how unusual a result at least as extreme as the observed result would be under a specified null model. The unadjusted p-value is 0.025. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of the difference.

The interval and p-value are unadjusted. The registry record does not report whether the primary analysis included covariate adjustment, how the original analysis handled missing values, or whether a multiplicity procedure was planned. These unknowns limit how far the recalculated statistics can stand in for the original analysis.

Why can over-dispersion matter?

Over-dispersion occurs when variability in the data is greater than a model assumes. In general, if a model underestimates variability, its standard errors can be too small and its confidence intervals too narrow. This concern is especially relevant to count outcomes, where variation can exceed the level expected under a simple count model. The primary endpoint is continuous, and the registry record does not describe a count model or report an over-dispersion assessment. There is no basis in the record to claim that over-dispersion affected this primary estimate.

What is the role of the standardized effect?

A standardized effect expresses a contrast relative to a measure of variability, rather than in the endpoint's original units. The recalculated standardized effect for the primary comparison is -0.679. It can help express the contrast on a scale that is less dependent on the original measurement units, but it does not replace the mg/dL estimate or its confidence interval. The pooled standard deviation is marked as approximate in the recalculation, so the standardized value should also be treated as an approximation.

What is not known about the original analysis?

The registry record does not report the original analysis model, the prespecified statistical plan, the handling of missing endpoint data, or any multiplicity adjustment. It also does not report a sponsor analysis for the primary endpoint. The posted recalculation is therefore useful as a transparent summary based on the posted arm statistics, but it should not be mistaken for a complete reconstruction of the trial's original analysis.

11. Confidence Intervals and Precision

The primary contrast is -80.64 mg/dL, with a 95% confidence interval from -151.2 to -10.09 mg/dL. The interval lies below zero, but its span indicates meaningful uncertainty about the magnitude of the difference. Values closer to zero and values representing a larger negative contrast are both compatible with the recalculated estimate under its assumptions. Precision is therefore different from whether the interval includes zero: an interval can exclude zero and still leave substantial uncertainty about effect size.

The arm-specific standard errors are 24.01 for pioglitazone and 26.82 for placebo. Standard errors summarize uncertainty in the estimated means, not the range of individual participant responses. The record does not provide individual-level outcome distributions or a model-based adjusted interval, so those features cannot be assessed from the posted summaries.

The interval is based on an unadjusted recalculation from posted summary statistics. Its interpretation relies on the calculation's assumptions, including the approximate pooled variability noted in the recalculation. The registry record does not report whether the sponsor's original analysis produced a different interval or used a different approach. The discrepancy flag is false, but that does not supply missing details of the original analysis.

Precision also depends on the endpoint population: the record describes U.S. participants with both baseline and follow-up triglyceride values. It does not establish how the estimate would change under another missing-data approach or in a broader population. The numerical interval is thus informative for the posted comparison, while leaving questions about model specification, missing observations, and generalizability unresolved.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total events. They illustrate sample sizes under a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is set as a stated fraction of the observed effect. The registry record does not report the original trial's sample-size assumptions or justification.

For planning at 100% of the observed effect, the illustration gives 35 participants per arm for 80% power and 46 participants per arm for 90% power. Planning for 75% of the observed effect gives 61 participants per arm for 80% power and 81 participants per arm for 90% power. Planning for 50% of the observed effect gives 137 participants per arm for 80% power and 183 participants per arm for 90% power.

In plain terms, the 80% power column describes the planned number per arm under which the calculation would have an 80% chance of detecting the specified effect if its assumptions held. The 90% power column uses a higher chance of detection and therefore requires a larger planned sample under the same effect assumptions. Smaller effects need more participants to detect, as reflected by the increasing figures when the assumed effect is reduced.

These are planning illustrations based on the observed effect, not observed power for the completed trial. They do not show that the trial was designed with these assumptions or that it had a particular probability of finding the posted result.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -80.640; pooled standard deviation σ = 118.713.
2. n = 2 × (1.96 + 0.842)² × 118.713² / -80.640² = 34.0, rounded up to 35 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- -80.64                      # observed difference in means
sigma <- 118.712628284803                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(231894)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 35 (80% power), 46 (90% power)
 75% of observed effect: n per arm = 61 (80% power), 81 (90% power)
 50% of observed effect: n per arm = 137 (80% power), 183 (90% power)
power.t.test: 35 
Simulated power at n = 35 per arm: 0.801 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when several hypotheses or outcomes are assessed. In general, as the number of statistical comparisons increases, the chance of at least one result appearing notable by chance can increase. This trial lists multiple secondary measures covering cognitive and metabolic outcomes in addition to the primary triglyceride endpoint. The registry record does not report a multiplicity adjustment, testing hierarchy, or other strategy for controlling the interpretation of multiple comparisons.

The primary p-value is unadjusted, and the recalculated secondary comparison is also unadjusted. Without a reported analysis plan, it is not possible to tell whether these comparisons were intended as confirmatory tests, exploratory analyses, or part of a prespecified sequence. The unadjusted secondary results therefore warrant particular caution and should not be treated as a set of independent confirmations.

Missing data are also important. The primary endpoint description identifies participants with triglyceride values at baseline and three-month follow-up, indicating that the reported endpoint analysis is based on participants with those measurements. The record does not report why measurements were unavailable for other participants, how missing values were handled, or whether any imputation or sensitivity analysis was performed. It does not provide enough detail to assess whether the observed endpoint population differs systematically from participants without follow-up values.

The analysis population for outcomes other than the primary endpoint is not fully described. The registry record does not report whether analyses followed intention-to-treat principles, how protocol deviations were handled, or whether all randomized participants were included in each analysis. These omissions do not establish that the analyses were biased, but they restrict assessment of robustness and generalizability. The original statistical plan and missing-data methods are not reported in the registry record.

15. 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 3% reporting threshold.

Most frequent non-serious adverse event terms

For serious adverse events, the registry record lists 0 events among 31 participants at risk in the pioglitazone group and 0 among 25 at risk in the placebo group. This means no serious adverse events are recorded for either group in the posted figures. It does not establish that the risks are identical or that no events could occur outside the recorded observation.

For other adverse events, the record lists 1 event among 31 at risk in the pioglitazone group and 2 among 25 at risk in the placebo group. These are event counts, not necessarily counts of participants experiencing at least one event. The record does not provide the event terms in the serious-adverse-event list, and it does not report death counts.

Small event counts are difficult to compare precisely. The registry figures alone do not describe event severity beyond the serious versus other classification, timing, duration, or causal attribution. They should be read as the safety information recorded for this trial, not as a complete estimate of risk in other populations or settings.

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

The posted primary estimate describes a lower average change in serum triglycerides in the pioglitazone arm than in the placebo arm among the U.S. participants with baseline and three-month values. The recalculated contrast is -80.64 mg/dL, and its 95% confidence interval ranges from -151.2 to -10.09 mg/dL. This is evidence about one measured outcome over the specified follow-up, under an unadjusted calculation from posted summaries.

The result does not establish the clinical importance of the difference. The registry record does not provide a threshold for a clinically meaningful change, and statistical uncertainty is substantial relative to the point estimate. It also does not show whether any difference persisted beyond the follow-up, affected other health outcomes, or applied to people outside the endpoint population.

The trial compared pioglitazone plus lifestyle diet with placebo plus lifestyle diet. Its contrast therefore does not isolate lifestyle diet, and it does not compare the intervention with an absence of lifestyle measures. The trial's focus included glucose, lipid, and cognitive measures, but the primary result alone cannot answer questions about all of those domains.

The posted secondary information is incomplete, and the registry record does not report the full original analysis plan, handling of missing outcomes, or multiplicity procedures. These gaps matter when assessing how much weight to give individual findings. The results are best understood as estimates from a particular randomized comparison and analysis population, not as medical advice or a universal conclusion about treatment.

17. Why This Trial Matters Statistically

This trial illustrates several statistical issues that arise in clinical research. First, a continuous endpoint can be summarized as an average change and an arm-to-arm contrast, but the estimate needs its uncertainty interval to show how precisely the difference is known. The primary comparison includes both a point estimate and a confidence interval, allowing readers to consider magnitude as well as compatibility with no difference.

Second, the endpoint-specific population matters. The primary estimate is based on U.S. participants with baseline and follow-up triglyceride measurements, while the overall enrollment figure describes the trial more broadly. Keeping those populations distinct prevents the analysis result from being mistaken for an estimate from every enrolled participant.

Third, the posted estimate is an unadjusted recalculation from summary statistics, while the record does not describe the original model. This distinction is important: summary-based recalculations can aid interpretation but cannot recover undocumented modeling choices or the treatment of missing data.

Finally, the trial's range of secondary measures raises questions about multiplicity and selective availability of results. A primary outcome and multiple secondary outcomes should not automatically be given equal evidential weight. This record shows why outcome definitions, populations, uncertainty, and reporting of analysis methods are all needed to interpret clinical trial results responsibly.

18. 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 endpoint is restricted to the U.S. sample with triglyceride measurements at baseline and three-month follow-up. The registry record does not explain why other participants may not have contributed those measurements or report methods for handling missing data. This makes it difficult to assess whether the endpoint population represents the randomized groups as a whole.

The record labels the arm values as least-squares means but does not describe the model, covariates, or adjustment procedure. The arm-to-arm estimate is an unadjusted recalculation from posted summaries, rather than a reported sponsor analysis. Although the discrepancy flag is false, the original analysis details remain unreported.

The registry lists several secondary outcomes, but results are not detailed for all of them. It does not report how multiplicity was addressed or whether the outcome hierarchy was prespecified. The safety figures are sparse and do not include reported death counts, detailed event terms, or information about event timing and attribution.

The follow-up described for the primary endpoint is 3 months, so the posted comparison does not address longer-term outcomes. The registry record also does not provide a detailed rationale for the primary endpoint, the original sample-size justification, randomization procedures, or masking procedures. These limitations affect how fully the trial methods and the broader relevance of its estimates can be evaluated.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Pioglitazone as a Treatment for Lipid and Glucose Abnormalities In Patients With Schizophrenia — NCT00231894.
• Registry record last updated 2017-12-11; status last verified 2017-11.
• Results first posted 2017-12-11.
• 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.

22. Record Summary

The registry record describes a completed, phase 4 randomized, quadruple-masked, parallel-group trial with enrollment of 56. It compared pioglitazone plus lifestyle diet with placebo plus lifestyle diet in people with schizophrenia receiving antipsychotic medication. The primary endpoint was change from baseline in serum triglycerides at 3 months among participants at U.S. sites with both measurements.

The posted primary arm means were -49.65 mg/dL for pioglitazone and 30.99 mg/dL for placebo. The unadjusted recalculated contrast was -80.64 mg/dL, with a 95% confidence interval from -151.2 to -10.09 mg/dL. The registry record does not report the original statistical model, missing-data methods, or multiplicity procedures. The primary result should therefore be interpreted within the endpoint population and the limits of the posted analysis details.