Effects of Pravastatin on Cholesterol, Inflammation and Cognition in Schizophrenia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effects of Pravastatin on Cholesterol, Inflammation and Cognition in Schizophrenia (NCT01082588), a randomized phase 4 study. The primary outcome is Change in LDL-cholesterol Between Baseline and Week 12.

2. Clinical Question

The trial asked whether pravastatin could affect cholesterol, inflammation, and cognition in people with schizophrenia or related disorders who were taking antipsychotic medication. The primary question is narrower: how did mean LDL-cholesterol change from baseline to week 12 in the pravastatin group compared with the placebo group?

The registry’s description also gives broader aims concerning inflammation and cognition, but the posted primary endpoint is LDL-cholesterol change. The record lists 7 outcome measures and no posted secondary analyses. It therefore does not provide a basis for conclusions about the other aims. The central statistical comparison is the difference in average change between the two groups, together with its confidence interval. That comparison describes the posted data; it does not, by itself, establish effects on other outcomes or broader clinical consequences.

3. Trial Design

The registry describes a completed Phase 4 study with randomized allocation, quadruple masking, and a parallel-group model. Participants were assigned to pravastatin or placebo, with both groups receiving their assigned daily treatment for 12 consecutive weeks. The registry lists an enrollment of 60. The study start is recorded as 2010-06 and primary completion as 2012-09.

The outcome table reports 23 participants in the pravastatin group and 23 in the placebo group for the primary LDL-cholesterol comparison. The safety section, in contrast, lists 30 participants at risk in each group. These are different reported counts for different sections of the record. The registry does not explain the relationship between them, so they should not be treated as interchangeable or used to infer the reason for the difference.

Randomization is intended, in general, to make treatment groups comparable on average, including for factors that are not measured. Masking is intended to reduce the influence of treatment knowledge on participants’ experience and on study procedures or assessments. The record identifies quadruple masking but does not specify which people or roles were masked. The parallel model means the record describes separate groups receiving pravastatin or placebo, rather than a design in which participants switch between treatments.

The registry record does not report the original sample-size justification, the detailed allocation procedure, or the trial’s complete statistical analysis plan. It also does not report interim analyses. Those details cannot be reconstructed from the posted outcome figures or the planning illustration.

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

Analysis population for the primary outcome, as defined in the registry: One participant from the pravastatin group and two from the placebo group had triglyceride levels above 400. Per Massachusetts General Hospital Laboratories policy, LDL is not run as a part of the complete metabolic panel (CMP) when triglycerides are above 400, and therefore could not be included in the final analysis.

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 in LDL-cholesterol Between Baseline and Week 12, measured over Baseline, week 12.

The primary endpoint is the change in LDL cholesterol between baseline and week 12, expressed in mg/dl. LDL cholesterol is a continuous measurement, so the registry reports a mean change and a standard deviation for each group. The comparison then summarizes how far apart the two group means were.

Change from baseline is used to describe how a measurement differs at follow-up from its starting value. Comparing changes between randomized groups can help distinguish the pattern in the treatment group from change over the same period in a placebo group. In general, using a continuous measure retains information about the size of individual measurements rather than reducing them to a category. The endpoint is specifically about LDL cholesterol; it is not a direct measure of inflammation, cognition, or clinical events.

The registry reports that one participant in the pravastatin group and two in the placebo group had triglyceride levels above 400. It states that, under Massachusetts General Hospital Laboratories policy, LDL is not run as part of the complete metabolic panel when triglycerides are above 400, so those measurements could not be included. The record does not describe other missing-data handling, or clarify whether any additional participants were excluded from this endpoint for other reasons.

7. Results Posted in the Registry

• Pravastatin: mean -25.57 versus -2.91 on control; difference -22.65 (95% CI -37.45 to -7.86), standardized effect d = -0.88; z-test p = 0.003.

8. Statistical Interpretation

9. Statistical Methods Explained

The posted primary endpoint is continuous: it records LDL-cholesterol change on a measurement scale, and the registry reports group means. The primary comparison is a mean difference, specifically -22.65 mg/dl. A mean difference keeps the result in the endpoint’s original units, making it possible to describe the estimated separation between groups without translating it into a different scale.

What the group summaries describe

The mean is a summary of the central value in a group. The standard deviation describes how spread out measurements are around that group’s mean. For the pravastatin group, the reported mean change is -25.57 mg/dl and the standard deviation is 32.26 mg/dl. For placebo, the mean change is -2.913 mg/dl and the standard deviation is 16.43 mg/dl. A standard deviation describes variability among observations; it is not the standard error, and it is not a confidence interval for the mean.

The registry also reports a pooled standard deviation of 25.6 mg/dl and a standardized effect estimate of -0.885. In general, a standardized effect expresses a difference relative to a measure of variability, which can help describe the size of a difference without using the original units. It is less directly interpretable in clinical measurement units than the mean difference. The record does not provide a full methods description for how the pooled estimate or standardized effect was calculated.

How to read the comparison

The mean difference is reported with a 95% confidence interval from -37.45 to -7.855 mg/dl. A confidence interval describes a range of values compatible with the data and the interval method’s assumptions. It conveys uncertainty in the estimated between-group difference, not the spread of individual participants’ measurements. The registry does not identify the precise interval procedure or statistical test used. The posted p-value is 0.003, but the record does not provide enough detail to verify the analysis method independently.

In general, a p-value is calculated under a null hypothesis, often a hypothesis of no between-group difference. Its interpretation depends on the statistical model and test assumptions. It is not a measure of the size of the effect, and it should not be used alone to judge precision or importance. The estimate and confidence interval are needed to see the direction and range of the reported difference. Here, the interval’s endpoints are both negative; this is consistent with the direction of the posted mean difference.

Variability and analysis choices

When measurements vary more among participants, estimating a group mean or a difference in means is generally less precise if other aspects of the study are unchanged. Unequal group variability can also make analysis choices important. In general, statistical methods may account for unequal variances, adjust for baseline measurements, or use other models. The registry record does not report whether such approaches were used for this endpoint. It would therefore be inappropriate to assign a particular test or adjustment to this result based only on the summaries.

In general, over-dispersion means that observed variability is greater than a model expects. When a model fails to account for such variability, its uncertainty estimates can be too small. That concept is relevant to choosing an analysis for some kinds of data, but this record does not report an over-dispersion assessment or a model that would allow one to evaluate it. The primary outcome is reported as a continuous measure, and the available summary statistics alone do not establish whether any particular distributional assumption was checked.

Planning illustrations are not trial-design documentation

The sample-size table is a planning illustration based on the observed effect, with two-sided significance level 0.05, equal allocation, and a normal approximation. It shows hypothetical participant counts per arm for detecting stated fractions of the observed effect at the listed power levels. This is not the same as the trial’s original design calculation. The original sample-size justification and its assumptions are not reported in the registry record.

In general, planning calculations ask how many participants might be needed to detect an effect under specified assumptions. Their results depend on the target effect, variability, significance level, power, allocation, and analysis method. A planning illustration based on the observed effect is especially dependent on that estimate and should not be mistaken for evidence about the probability that the completed trial detected an effect. It also cannot explain why the registry reports different participant counts in the primary outcome and safety sections.

What is not specified

The registry reports no posted analyses beyond the primary comparison and no posted secondary analyses. It does not provide the full statistical analysis plan, the analysis population definition, any interim-analysis plan, or a detailed missing-data strategy. It is not reported whether the comparison was adjusted for baseline values or other covariates, whether the reported p-value accounts for multiple outcomes, or what exact procedures were used for the confidence interval. Those omissions matter because analysis decisions affect how estimates and uncertainty are interpreted. The posted figures can be described as recorded, but they do not support claims about unreported methods.

10. Confidence Intervals and Precision

The 95% confidence interval for the mean difference runs from -37.45 to -7.855 mg/dl. It expresses uncertainty about the between-group estimate, under the interval method’s assumptions. The interval does not describe where most individual participants’ changes fell; the group standard deviations are the summaries of participant-level spread.

Both endpoints of the reported interval are below zero, as is the estimated mean difference of -22.65 mg/dl. Thus, the range displayed by this interval is in the direction of a more negative mean change in the pravastatin group relative to placebo. The interval is still a range rather than a single established value. Its width reflects uncertainty in the comparison, and the precise interpretation depends on the statistical method, which is not described in detail in the registry record.

The confidence interval should be considered with the primary outcome counts of 23 participants in each group and with the reported group variability: standard deviations of 32.26 mg/dl and 16.43 mg/dl. These summaries indicate why the estimated contrast should not be read as an exact value applying to every participant. The record also notes that LDL measurements could not be included for one participant in the pravastatin group and two in the placebo group because of triglyceride levels above 400 and laboratory policy. It does not report a complete account of missingness or the method used to address it.

The interval does not capture every source of uncertainty. It does not account automatically for unreported analysis choices, possible multiplicity, or limitations in how participants with unavailable measurements were handled. Those issues cannot be quantified from the registry record.

11. 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 sample-size table and curve are planning illustrations, not a record of the study’s observed enrollment or a reconstruction of its original design. The figures are participants per arm, not total events. They use the observed effect as a reference and show planning for 100%, 75%, and 50% of that effect. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The original trial’s own sample-size assumptions and justification are not reported in the registry record.

The 80% power column describes the number of participants per arm that the planning calculation gives for an 80% chance of detecting the specified effect under its assumptions. The 90% power column gives the corresponding planning count for a 90% chance. These are conditional design calculations, not guarantees about what a study will find. The table shows 21 and 27 participants per arm for planning around 100% of the observed effect; 36 and 48 for planning around 75% of the observed effect; and 81 and 108 for planning around 50% of the observed effect, respectively for the 80% and 90% columns.

Smaller effects require more participants under these planning assumptions, because a smaller difference is harder to distinguish from variability. The table is not observed power. It should not be used to claim that the completed trial had a particular probability of detecting its result. The counts are illustrations based on the observed estimate, rather than evidence that the trial was originally designed with these targets.

12. 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 = -22.652; pooled standard deviation σ = 25.601.
2. n = 2 × (1.96 + 0.842)² × 25.601² / -22.652² = 20.1, rounded up to 21 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 <- -22.652                      # observed difference in means
sigma <- 25.6012546274592                     # 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(1082588)
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 = 21 (80% power), 27 (90% power)
 75% of observed effect: n per arm = 36 (80% power), 48 (90% power)
 50% of observed effect: n per arm = 81 (80% power), 108 (90% power)
power.t.test: 21.1 
Simulated power at n = 21 per arm: 0.799 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists 7 outcome measures, but no secondary analyses are posted. The LDL-cholesterol comparison is presented as the primary result. The record does not describe whether the trial assessed multiple outcomes in a prespecified hierarchy, whether any adjustment for multiple testing was planned or applied, or how the reported p-value should be interpreted in relation to the full set of outcomes. Without that information, the posted p-value should be treated as a result for the reported comparison, not as a summary of the whole family of study questions.

The record does not provide the full statistical analysis plan, an analysis population definition, or interim-analysis information. It also does not state whether the primary comparison used an intention-to-treat population, a per-protocol population, or another set of participants. The analysis population label and the rules for including participants are not reported in the registry record.

Some LDL measurements were unavailable because the registry states that laboratory policy did not run LDL as part of the complete metabolic panel when triglycerides were above 400. This applied to one participant in the pravastatin group and two in the placebo group. The record does not explain whether these were the only missing primary measurements, how missing observations were handled in the analysis, or whether any imputation or sensitivity analyses were used.

The primary outcome table lists 23 participants in each group, while the safety section lists 30 at risk in each group and the enrollment field reports 60. The registry does not reconcile those counts. It is not possible to determine from the record why the counts differ or whether they reflect different analysis populations, measurement availability, or other factors. These reporting gaps limit evaluation of completeness and of the extent to which the posted comparison represents all enrolled participants.

14. 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 0% reporting threshold.

Most frequent non-serious adverse event terms

The safety section lists 30 participants at risk in each group. It reports no serious adverse events in either group. It also lists 5 other adverse events in the pravastatin group and 1 in the placebo group. These are the counts posted in the record; they should not be confused with the primary endpoint analysis counts of 23 per group.

The registry does not list serious adverse-event terms, and death counts are not reported in the registry record. It does not provide event descriptions, severity categories for the other adverse events, timing, or information about how many participants experienced more than one event. The record also does not specify the definitions or collection procedures used for these categories.

These figures are a limited summary of reported events. They do not establish that no other events occurred, and they do not provide enough detail to compare event rates or assess causes. The different counts for safety and the LDL outcome are reported as such; the registry does not explain their relationship.

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

The posted primary result is a comparison of mean LDL-cholesterol change from baseline to week 12. The pravastatin group’s mean change was -25.57 mg/dl, and the placebo group’s mean change was -2.913 mg/dl. The registry reports a mean difference of -22.65 mg/dl, with a 95% confidence interval from -37.45 to -7.855 mg/dl and a p-value of 0.003. These results describe the endpoint and participants represented in the posted comparison.

They do not, on their own, establish effects on inflammation or cognition, even though those topics appear among the study’s aims. The record posts no secondary analyses. The LDL result also does not demonstrate an effect on heart disease, diabetes, brain function, or other clinical outcomes. Those are not the primary measure, and no such results are included in the posted primary comparison.

The result should be interpreted in light of the reported counts and variability. The primary outcome comparison includes 23 participants per group, with group standard deviations of 32.26 mg/dl and 16.43 mg/dl. The registry reports an enrollment of 60 and safety counts of 30 per group, but does not reconcile these figures. It also describes LDL measurements that could not be included for participants with triglyceride levels above 400, without detailing all missing-data handling.

The confidence interval conveys uncertainty around the estimated group difference, but does not address unreported analysis decisions or the absence of secondary results. The registry record does not give enough detail to establish the analysis population, the exact statistical test, or whether multiplicity was considered. The posted data are therefore best understood as a specific estimate for the reported LDL endpoint, rather than a complete account of all study aims or all possible consequences of treatment.

16. Why This Trial Matters Statistically

This trial offers a useful example of why an endpoint, an estimate, and its uncertainty should be read together. A group mean change describes what was observed within each group, while the mean difference directly compares those changes. The confidence interval adds information about precision that the point estimate alone cannot provide. The standard deviations, in turn, describe variation among participants and should not be mistaken for uncertainty in the between-group estimate.

The record also illustrates that different sections can report different participant counts. The primary LDL comparison has 23 participants per group, while the safety section lists 30 per group and the enrollment field lists 60. Those counts answer different reporting questions and are not reconciled by the registry. A careful reading keeps them distinct rather than treating any one count as the denominator for every result.

Finally, the study summary names several aims, but the posted analysis is limited to the primary LDL result. The absence of posted secondary analyses means that the broader aims cannot be evaluated from this record. The planning table is also distinct from the original trial design: it illustrates hypothetical sample sizes under stated assumptions but does not document how the completed trial was planned. Together, these distinctions show how statistical interpretation depends not just on the numerical result but on the endpoint, analysis population, uncertainty, and completeness of reporting.

17. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The registry provides a primary endpoint comparison but does not report a complete statistical analysis plan. The exact test and confidence-interval procedure, any baseline or covariate adjustment, the analysis population definition, and missing-data methods are not reported in the registry record. The registry also does not reconcile the enrollment of 60 with the 23 participants per group in the LDL analysis and the 30 per group listed for safety.

The endpoint description identifies LDL measurements that could not be included for one participant in the pravastatin group and two in the placebo group because triglyceride levels were above 400 and laboratory policy did not run LDL in that circumstance. Other reasons for unavailable data, if any, and the handling of missing measurements are not described. The record gives no posted secondary analyses despite listing 7 outcome measures, so the broader study aims cannot be assessed using these results.

The registry identifies quadruple masking but does not specify which roles were masked. It does not report the original sample-size justification or interim analyses. The safety summary reports no serious adverse events and gives counts for other events, but does not provide event terms, definitions, or detailed descriptions. These reporting limits do not change the posted numbers, but they constrain what can be concluded from them.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Effects of Pravastatin on Cholesterol, Inflammation and Cognition in Schizophrenia — NCT01082588.
• Registry record last updated 2014-04-28; status last verified 2014-03.
• Results first posted 2014-04-28.
• 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.

21. Record Summary

This completed Phase 4 randomized, quadruple-masked, parallel-group trial compared pravastatin with placebo in people with schizophrenia-spectrum disorders. The reported primary endpoint was mean change in LDL cholesterol from baseline to week 12. The registry lists an enrollment of 60 and records study start in 2010-06 and primary completion in 2012-09.

The primary outcome table reports 23 participants per group. Mean change was -25.57 mg/dl with pravastatin and -2.913 mg/dl with placebo; the mean difference was -22.65 mg/dl, with a 95% confidence interval from -37.45 to -7.855 mg/dl and a p-value of 0.003. The registry lists 7 outcome measures but posts no secondary analyses. Safety counts are reported separately, with 30 participants at risk in each group.

The registry record does not report the full analysis plan, original sample-size justification, analysis population definition, or comprehensive missing-data handling. It does not reconcile the different participant counts across sections.