A Study to Evaluate the Relative Bioavailability (BA) and Effect of Food on TAK-831 Tablet Formulations in Healthy Participants: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate the Relative Bioavailability (BA) and Effect of Food on TAK-831 Tablet Formulations in Healthy Participants (NCT03706469), a randomized phase 1 study. The primary outcome is AUClast: Area Under the Plasma Concentration-time Curve From Time 0 to the Time of the Last Quantifiable Concentration for TAK-831.

2. Clinical Question

The study was designed to assess relative bioavailability of the T3 tablet formulation compared with T2 under fasting conditions, and the effect of food on the pharmacokinetics of T3. These are pharmacokinetic questions about drug exposure in healthy participants, rather than questions about clinical benefit or treatment of a condition. The registry record also lists treatment-emergent adverse events as a secondary outcome.

3. Trial Design

This was a phase 1, open-label, randomized, single-dose, five-period, five-treatment study in healthy participants. The registry record lists an enrollment of 16 and a crossover model. In a crossover study, participants receive more than one treatment condition in an assigned sequence, so comparisons can be made within participants as well as between treatment conditions. The treatment sequences listed in the record vary the order of fasted T2 and T3 doses and include a fed T3 600 mg period.

The posted pharmacokinetic results cover five conditions: TAK-831 T2 50 mg fasted, TAK-831 T3 50 mg fasted, TAK-831 T2 600 mg fasted, TAK-831 T3 600 mg fasted, and TAK-831 T3 600 mg fed. The key comparisons are the two formulations at 50 mg under fasting conditions, the two formulations at 600 mg under fasting conditions, and T3 600 mg under fasted versus fed conditions.

Participants received a single oral dose in each treatment period. The primary endpoint was assessed from predose through multiple postdose time points up to 72 hours on Day 1. The registry record lists the study as completed, with a start date of 2018-10-18 and a primary completion date of 2018-12-11.

Open-label means masking was not used, as also indicated by the record. The specific randomization procedure, treatment-period washout details, and any measures used to assess period or carryover effects are not reported in the registry record. The stated study purpose is relative bioavailability and food-effect assessment; the record does not describe a clinical efficacy question.

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: Healthy participants were enrolled and randomized in one of the 4 treatment sequences to receive TAK-831.

Overall Study

Analysis population for the primary outcome, as defined in the registry: The pharmacokinetic (PK) analysis set included all participants from the safety set who had at least 1 measurable postdose TAK-831 plasma concentration.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: The safety analysis set included all participants who received at least one dose of the study drug(s).

6. Primary Endpoint

The registry lists the primary outcome as AUClast: Area Under the Plasma Concentration-time Curve From Time 0 to the Time of the Last Quantifiable Concentration for TAK-831, measured over Day 1 pre-dose and at multiple time points (up to 72 hours) post-dose.

The primary endpoint was AUClast for TAK-831: the area under the plasma concentration-time curve from time 0 to the time of the last quantifiable concentration. The record gives the unit as nanogram*hour per milliliter (ng*hr/mL). It was measured from predose through multiple time points up to 72 hours after dosing on Day 1.

AUClast summarizes plasma exposure over the observed period. It incorporates measured concentrations across time, rather than relying on a single concentration reading. In general, pharmacokinetic studies use exposure measures such as this to compare how formulations or conditions relate to the amount of drug measured in plasma over time. This makes the endpoint relevant to the stated relative-bioavailability and food-effect questions.

The reported arm values are geometric means, with geometric coefficients of variation describing variability. Geometric summaries are commonly used for exposure measures that are analyzed on a multiplicative scale. They describe the center and spread of the observed pharmacokinetic measurements; they do not, by themselves, establish a clinical effect. The registry record does not report an additional clinical outcome as the primary endpoint.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a geometric mean ratio represent?

A geometric mean ratio expresses one treatment condition relative to another on a multiplicative scale. A value of 100 indicates equal geometric means on the percentage scale used in the posted results. Values below or above 100 indicate that the numerator condition's geometric mean is numerically lower or higher than the comparator, respectively. The exact ordering of each pair matters when interpreting direction, so the comparisons should be read in the stated group order: T3 versus T2 for each fasted dose, and fed versus fasted for T3 600 mg.

This ratio scale is useful for pharmacokinetic measures because exposure values can be skewed and relative changes are often more interpretable than absolute differences. That is a general statistical rationale, not a statement about the specific analysis procedure used in this study. The registry record does not report the model or transformation used to estimate the ratios.

How should the confidence intervals be read?

A confidence interval describes the uncertainty associated with an estimated quantity under the assumptions of the analysis. In general, a 90% confidence interval is produced by a procedure designed to cover the true parameter in 90% of repeated samples under its statistical assumptions. It is not a probability statement that this particular interval contains the true value. Wider intervals indicate less precision than narrower intervals, all else equal.

The posted intervals are 74.50 to 93.80 for the 50 mg fasted formulation comparison, 91.10 to 114.70 for the 600 mg fasted formulation comparison, and 152.55 to 195.58 for the fed-versus-fasted T3 600 mg comparison. The record does not state the equivalence margins used to interpret the intervals. Without those bounds, the intervals should not be treated as a stand-alone determination of equivalence.

Why does the crossover design matter?

In a crossover design, participants receive multiple conditions in a planned sequence. General advantages can include comparisons that account for stable differences between participants, because each participant contributes information under more than one condition. However, valid interpretation can depend on features such as the period structure, treatment sequence, and potential carryover. The registry record lists randomized treatment sequences and five treatment periods, but detailed methods for handling period effects, sequence effects, or carryover are not reported in the registry record.

The arm summaries do not reveal the paired participant-level differences that are central to a crossover comparison. They are useful descriptive summaries, but they cannot reproduce a model-based crossover analysis without participant-level information or fuller statistical details. The registry record does not identify the estimation model used for the reported ratios.

What does the variability measure add?

The arm summaries report geometric coefficients of variation: 28.7, 37.0, 24.8, 41.8, and 29.0 for the five conditions, in the order listed in the registry record. A coefficient of variation expresses variability relative to the geometric mean, rather than in the original measurement units. In general, relative variability can help describe how dispersed exposure measurements are when values differ in scale. These values do not provide the participant-level distribution or the precision of the ratio estimates by themselves.

The primary endpoint's pharmacokinetic analysis set included participants from the safety set who had at least one measurable postdose TAK-831 plasma concentration. The number excluded from that analysis set, the handling of missing measurements, and the rules for nonquantifiable concentrations beyond the endpoint definition are not reported in the registry record.

What statistical details remain unclear?

The record lists three reported ratio analyses and identifies them as equivalence analyses. It does not report the estimation method, model terms, equivalence margins, hypothesis-testing procedure, or adjustment for multiple comparisons. It also does not report whether any interim analyses were planned or conducted. The absence of those details limits independent assessment of how the reported intervals were obtained and how the analyses were intended to support formal decisions.

General statistical interpretation should keep the distinctions clear: geometric means describe condition-level exposure; ratios describe relative exposure; confidence intervals describe uncertainty under an analysis; and an equivalence conclusion requires prespecified bounds and an appropriate analysis. The posted summaries are informative about the observed pharmacokinetic pattern, but do not supply every detail needed to reproduce or fully audit the sponsor's analysis.

10. Analyses Reported by the Sponsor

The registry record reports geometric mean ratios and 90% confidence intervals for the formulation and food-condition comparisons. These are the reported analyses to use when describing the sponsor's posted results. The record labels them as equivalence analyses, but does not list the equivalence bounds or the statistical method used to estimate the ratios.

The arm-level geometric means and coefficients of variation are descriptive summaries. A simple unadjusted recalculation from those summaries would not reproduce a crossover analysis that uses participant-level paired information or model adjustments. No recalculated ratio estimates are included in the registry record, and no discrepancy is flagged between reported and recalculated analyses. The record also does not provide enough method detail to independently reproduce the sponsor's estimates from the arm summaries alone.

Accordingly, the reported ratios and intervals should be distinguished from any calculation based only on summary values. The confidence intervals provide the reported measure of uncertainty, while the missing equivalence margins prevent a conclusion about whether a specified formal equivalence criterion was met.

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists one primary pharmacokinetic endpoint and a secondary outcome concerning the percentage of participants who experience at least one treatment-emergent adverse event. It also reports three pairwise pharmacokinetic comparisons: T3 versus T2 at 50 mg under fasting conditions, T3 versus T2 at 600 mg under fasting conditions, and T3 600 mg under fed versus fasting conditions.

How multiplicity was handled is not reported in the registry record. In general, when several comparisons are examined, the chance of at least one false-positive finding can increase if each comparison is assessed separately without adjustment. That general concern does not establish that any particular adjustment was required or used in this study. The record does not state whether the comparisons were prespecified as a family, whether any hierarchy was used, or whether confidence intervals or tests were adjusted.

The analysis population is described for the primary pharmacokinetic endpoint: it included participants in the safety set with at least one measurable postdose TAK-831 plasma concentration. The registry record does not report the number in that analysis set separately, the amount or pattern of missing pharmacokinetic data, or the method used to handle missing values. It does not describe exclusions, protocol deviations, or sensitivity analyses for the primary endpoint.

The record's stated analysis population is useful context, but it is not enough to determine whether the analysis followed an intention-to-treat, per-protocol, or other framework. Such classifications and the rules used to define them are not reported in the registry record. The record also does not report whether interim analyses took place or describe a statistical analysis plan. These omissions limit assessment of prespecification and multiplicity control.

12. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent non-serious adverse event terms

The safety results are reported by treatment condition. The registry record lists zero serious adverse events in each of the five conditions and zero deaths in each condition. For non-serious adverse events, it lists 0 affected participants for T2 50 mg fasted, 2 for T3 50 mg fasted, 0 for T2 600 mg fasted, 1 for T3 600 mg fasted, and 0 for T3 600 mg fed. Each condition lists 16 participants at risk for these summaries.

These are descriptive counts. The record does not report statistical comparisons of adverse-event frequencies between conditions, nor does it list serious adverse-event terms. In a crossover study, participants may contribute to more than one treatment condition, so condition-specific counts should not be added to infer a count of unique participants. The registry record does not provide participant-level information to assess overlap across treatment periods.

Zero reported serious adverse events in these condition summaries means no such events are listed for the study conditions in the posted results; it does not establish absence of risk. The study is small and focused on pharmacokinetics, and the safety counts should be interpreted within that limited descriptive context.

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

The posted results describe TAK-831 exposure across two tablet formulations, two dose levels under fasting conditions, and a fed-versus-fasted comparison for the T3 formulation. The reported ratios indicate different numerical patterns across these comparisons: the 50 mg formulation ratio is below 100, the 600 mg formulation ratio is close to 100, and the food-condition ratio is above 100. The confidence intervals show the precision reported for those estimates, with the limits differing across comparisons.

These findings concern AUClast, a plasma exposure measure, in healthy participants. They do not establish clinical benefit, lack of benefit, or effectiveness for any medical condition. The study was not described as an efficacy trial, and the registry record lists healthy volunteers as the condition.

The record labels the ratios as equivalence analyses, but the equivalence margins are not reported in the registry record. A numerical ratio near 100 alone does not demonstrate equivalence; that judgment depends on prespecified bounds and a suitable analysis. Similarly, the food-condition ratio describes the posted exposure comparison but does not, by itself, establish the clinical implications of food-related differences.

The study's crossover design is relevant because each participant could receive multiple conditions, but the posted aggregate results do not reveal participant-level paired responses. The reported summaries should therefore be read as group-level results rather than a full account of individual variability or treatment sequence effects. The missing analysis details also limit independent reconstruction of the estimates.

The adverse-event counts are descriptive and do not support comparative safety conclusions. Overall, the registry record offers a focused account of relative pharmacokinetic exposure and reported event counts, while leaving important details about analysis methods, equivalence criteria, missing data, and multiplicity unreported.

14. Why This Trial Matters Statistically

This study illustrates why pharmacokinetic questions and clinical outcome questions should be kept distinct. AUClast is an exposure measure that can compare formulations and dosing conditions; it is not itself a measure of symptom improvement, disease control, or patient benefit. The posted comparisons therefore address the stated bioavailability and food-effect aims, not clinical efficacy.

It also demonstrates the value of reporting ratios with confidence intervals rather than relying only on separate condition means. Ratios describe relative exposure directly, while intervals communicate uncertainty. The results have different numerical patterns across the dose and food comparisons, making it important to interpret each comparison on its own terms rather than summarize all conditions with one broad statement.

The crossover design provides a setting for within-participant comparisons, but aggregate summaries do not show the paired responses that underlie such analyses. Detailed reporting of the model, sequence and period handling, missing data, and equivalence margins would make the results easier to evaluate and reproduce. The registry record does not report those details.

Finally, the study highlights that a label such as equivalence analysis is not a substitute for stating the decision criteria. A ratio and confidence interval can be described precisely, but a formal conclusion requires knowing the bounds against which the interval was assessed. This distinction helps readers separate what the posted estimates show from what cannot be determined from the registry record.

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 registry record does not report the statistical model used for the crossover analyses, how period or sequence effects were handled, or whether carryover was assessed. It also does not provide the equivalence bounds, despite labeling the reported comparisons as equivalence analyses. These omissions limit evaluation of the formal interpretation and make it difficult to reproduce the estimates from the posted arm summaries.

Participant-level pharmacokinetic measurements and paired treatment contrasts are not reported in the registry record. Consequently, the condition-level means and variability measures do not show how consistent individual responses were across conditions. The missing-data rules, number of participants in the primary analysis set, and sensitivity analyses are also not reported.

The study was open-label and conducted in healthy participants. The registry record does not describe masking, and its stated population does not support extending these pharmacokinetic summaries to people with a specific illness. The record also does not report a clinical efficacy endpoint. The adverse-event counts are descriptive, with no comparative safety analysis listed.

Finally, the registry record does not state how multiplicity was addressed across the reported comparisons or provide a detailed statistical analysis plan. These limitations do not negate the reported summaries, but they constrain the conclusions that can be drawn from them and the extent to which an independent reader can assess the analysis.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study to Evaluate the Relative Bioavailability (BA) and Effect of Food on TAK-831 Tablet Formulations in Healthy Participants — NCT03706469.
• Registry record last updated 2021-06-14; status last verified 2021-06.
• Results first posted 2019-12-23.
• 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

This completed phase 1 study was randomized, open-label, and used a single-dose, five-period crossover design in healthy participants. The stated purpose was to compare the relative bioavailability of TAK-831 T3 and T2 under fasting conditions and assess the effect of food on T3 pharmacokinetics.

The primary endpoint was AUClast, summarized by geometric means and geometric coefficients of variation. The registry record reports geometric mean ratios with 90% confidence intervals for the two fasted formulation comparisons and the T3 fed-versus-fasted comparison. It also lists treatment-emergent adverse-event counts by condition, with zero serious adverse events and zero deaths reported across the conditions.

Important analysis details, including the equivalence margins, statistical model, missing-data handling, and multiplicity approach, are not reported in the registry record. The posted findings describe pharmacokinetic exposure and do not establish clinical benefit or a comparative safety conclusion.