Study of LY2409021 in Participants With Type 1 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of LY2409021 in Participants With Type 1 Diabetes (NCT01640834), a randomized phase 1 study. The primary outcome is Pharmacodynamics: Change From Baseline to Day 2 in 24-hour Insulin Dose.

2. Clinical Question

3. Trial Design

This was a randomized, double-masked, parallel-group Phase 1 study of LY2409021 in participants with type 1 diabetes. The registry record lists the study as completed and records enrollment of 20. The study began in 2012-07, with primary completion in 2012-09. These dates describe the registry record; they do not establish the duration of participation for every individual.

Three groups are described: 100 mg LY2409021, 300 mg LY2409021, and placebo. The assigned oral treatment was a single dose on Day 2. In the 100 mg group, one LY2409021 capsule and two placebo capsules were given. In the 300 mg group, three LY2409021 capsules were given. The placebo group received three placebo capsules. Each arm description also specifies 1 mg glucagon by intramuscular injection on Day 3.

The study summary describes a 7-day hospitalization period at a clinical research unit, screening within 30 days of study start, and telephone consultations within 5 days after discharge. The registry record does not report further details about the randomization procedure, allocation concealment, or which individuals were masked. It also does not report how masking was maintained for the different capsule contents.

The parallel-group structure means participants were assigned to a group rather than being described as crossing between treatments. The record does not report a crossover design or a washout between assigned treatments. A secondary outcome does refer to insulin dose during a drug washout period, but details of that period are not reported in the registry record.

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: All participants who received a dose of LY2409021 or placebo and had evaluable 24-hour insulin data.

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: All randomized participants.

6. Primary Endpoint

The registry lists the primary outcome as Pharmacodynamics: Change From Baseline to Day 2 in 24-hour Insulin Dose, measured over Baseline (Day 1), Day 2.

The primary endpoint was the mean absolute change in total insulin dose over 24 hours from Day 1 to Day 2, measured in insulin units. The registry defines the change as the Day 2 24-hour insulin dose minus the Day 1 24-hour insulin dose. It reports a group mean and standard deviation. This is a continuous outcome: individual changes can take different values across a range, rather than being limited to categories such as event or no event.

Insulin dose is the study’s direct measure of the amount of insulin needed over the specified 24-hour period. Comparing the change from baseline focuses on how the measured dose shifted between the two study days, rather than comparing only the Day 2 values. A change score can account for each participant’s starting dose in the definition of the outcome, although it does not by itself adjust for other participant characteristics or remove measurement variability.

The endpoint population included participants who received LY2409021 or placebo and had evaluable 24-hour insulin data. The record does not report how many otherwise eligible participants lacked evaluable data or how such data were handled. The planned observation window was short, and the primary endpoint concerns the change between baseline on Day 1 and Day 2. It therefore describes insulin dose over that study interval, not longer-term dose patterns or outcomes beyond the measured period.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous, so it is summarized with a mean and standard deviation. The mean describes the average change in insulin dose within a group. The standard deviation describes how spread out individual changes are around that group mean. These answer different questions: a small standard deviation indicates less observed variation, while precision in the mean estimate also depends on the amount of data and the variability.

What does the net mean difference represent?

A between-group mean difference compares the average outcome in one group with the average in another. The registry reports net mean differences for 100 mg and 300 mg LY2409021 against placebo. These are reported as -5.84 and -6.89 insulin units, respectively. The subtraction convention is not stated in the registry record, so the values can be quoted but their sign should not be translated into a treatment-versus-placebo direction without that definition. A group’s own mean change and a between-group difference are not interchangeable: the latter incorporates the comparison group’s observed mean.

How should the confidence intervals be read?

A confidence interval expresses uncertainty around an estimate under the assumptions of the statistical method. It gives a range of values that are compatible with the observed data and the method, rather than a guarantee that the true value lies within that particular interval. The reported intervals are 95% confidence intervals. Both span zero, which is the value representing no between-group mean difference. This means the intervals include values on either side of zero; it does not prove that the groups are identical.

Interval width matters as well as whether zero is included. Wider intervals indicate less precision about the difference. In a small study, estimates may be imprecise because there are relatively few observations and individual outcomes can have a substantial influence. The 300 mg comparison has a wider reported interval than the 100 mg comparison. The interval should therefore be considered alongside the point estimate, sample size, and observed variability, rather than treated as a binary verdict.

What does a one-sided t-test test?

The reported sponsor analyses used one-sided t-tests. In general, a one-sided test evaluates evidence in a specified direction, unlike a two-sided test, which considers departures in either direction. The registry record does not report the rationale for using a one-sided test or specify the directional hypothesis. The reported p-values, 0.054 for 100 mg and 0.083 for 300 mg, are conditional summaries under the test assumptions and the tested direction. They are not probabilities that the null hypothesis is true, nor measures of the size or clinical importance of a difference.

The registry also gives 95% confidence intervals alongside the one-sided p-values. These are different summaries and should be interpreted according to their stated form. The confidence intervals include zero, while the p-values come from one-sided tests. The record does not give additional detail on how the interval procedure relates to the one-sided tests. Readers should not assume the interval and test use identical inferential conventions merely because they are reported together.

Why does variability matter?

General statistical context: when outcomes vary substantially within groups, it becomes harder to estimate group means and differences precisely. A standard deviation is a measure of that observed spread; it is not a standard error or confidence interval. The reported standard deviations differ across the three groups. That pattern, together with the limited group sizes, is relevant when judging how much confidence to place in the point estimates.

General statistical context: over-dispersion occurs when observed variability is greater than a model assumes. It can make standard errors too small and confidence intervals too narrow if an analysis fails to accommodate the extra spread. The registry reports standard deviations for this continuous endpoint, but does not report a variance model or an assessment of over-dispersion. This issue is a general consideration, not a claim that over-dispersion was demonstrated in this study.

What do the planning illustrations mean?

The study’s planning illustrations are based on the observed effect and specify a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The assumed effect is a stated fraction of the observed effect. These are planning scenarios, not observed results or evidence that the completed trial had a particular probability of detecting an effect. The registry record does not report that these assumptions were the original trial’s design assumptions or its sample-size justification. The original sample-size justification is not reported in the registry record.

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("100 mg LY2409021", "300 mg LY2409021", "Placebo")
n   <- c(8, 8, 4)
m   <- c(1.85, 0.8, 7.69)
sp  <- c(7.15, 9.63, 2.99)
sd  <- sp      # posted standard deviations
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  100 mg LY2409021  8  1.850  7.150  -4.128  7.828
CARM  300 mg LY2409021  8  0.800  9.630  -7.251  8.851
CARM  Placebo  4  7.690  2.990  2.932  12.448

11. Analyses Reported by the Sponsor

The registry reports sponsor analyses comparing each LY2409021 dose group with placebo. For 100 mg, the reported net mean difference was -5.84 insulin units, with a 95% confidence interval of -11.82 to 0.14 and a one-sided t-test p-value of 0.054. For 300 mg, the reported difference was -6.89 insulin units, with a 95% confidence interval of -14.95 to 1.16 and a one-sided t-test p-value of 0.083. The registry does not state the subtraction convention for these net differences.

The registry record does not flag a discrepancy between the sponsor’s reported analyses and the posted group summaries. No separate recalculated numerical comparison is reported in the registry record, so there is no additional estimate to set against the sponsor’s values. The sponsor analyses should therefore be read as the reported inferential comparisons, not as an independently reproduced result with a second set of estimates.

These analyses use one-sided t-tests, while the accompanying intervals are stated as 95% confidence intervals. The record does not report further details about the testing direction or the interval procedure. Neither p-value alone establishes the magnitude or importance of a difference. The confidence intervals include zero, and their widths indicate that the estimates remain uncertain. The record does not report a sponsor conclusion that would add an interpretation beyond the numerical analyses.

12. Multiplicity, Missing Data and Analysis Populations

The primary outcome is one of eight posted outcome measures, and the registry lists eight analyses. Seven secondary outcomes are posted. They cover insulin dose on Day 2, pharmacokinetic measures of LY2409021, insulin dose during washout and to maintain euglycemia, and glucose responses following glucagon. The registry record does not report how the family of outcomes or comparisons was handled for multiplicity, including whether any adjustment was planned or whether the primary endpoint was treated as the sole confirmatory analysis.

General statistical context: when many hypotheses are tested, the chance of obtaining at least one apparently unusual result can increase. Multiplicity procedures can address that issue, but the choice depends on the study’s goals and analysis plan. Since a multiplicity strategy is not reported in the registry record, the sponsor-reported p-values should be interpreted as the stated individual tests, not as results known to have been adjusted across the outcome set.

The primary endpoint population is defined as participants who received a dose of LY2409021 or placebo and had evaluable 24-hour insulin data. The registry reports group sizes for the endpoint, but does not describe how many screened or enrolled participants lacked evaluable data, reasons for missingness, or imputation procedures. Handling of missing primary endpoint data is not reported in the registry record. It is also not reported whether the analysis included all randomized participants or only those meeting the endpoint population definition.

Missing outcomes can affect interpretation if the likelihood of missingness is related to treatment assignment or to the outcome itself. This is a general concern, not evidence that such a pattern occurred in this study. Without details about missingness and its handling, the posted analysis cannot show how robust the estimates would be to alternative assumptions about unavailable measurements.

The registry does not report an interim analysis plan, stopping rules, or a statistical analysis plan. The rationale for the one-sided tests, the prespecified direction of testing, and any planned subgroup analyses are not reported in the registry record. These omissions limit assessment of how closely the reported analyses follow an analysis plan established before results were known.

13. 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 reports zero participants with a serious adverse event in each of the three groups: 100 mg LY2409021, 300 mg LY2409021, and placebo. The numbers at risk for these summaries were 8, 8, and 4, respectively. No serious adverse event terms are listed. These are the posted serious-event counts during the reporting period; they do not show that the probability of a serious event is zero.

The record also reports other adverse events affecting 6 participants in the 100 mg group, 6 in the 300 mg group, and 4 in the placebo group. The record does not provide the event terms, severity, timing, or attribution in the adverse-event details listed for this analysis. Death counts are not reported in the registry record.

In general, an absence of reported serious events in a small study does not exclude uncommon harms or events that occur outside the observed period. These safety figures should be understood as descriptive counts from this study, not as a comparative safety estimate with demonstrated precision. The registry record does not report a formal statistical comparison of adverse events between groups.

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

The posted results describe a randomized comparison of two single LY2409021 doses with placebo, using the change in 24-hour insulin dose from Day 1 to Day 2 as the primary endpoint. The group means and sponsor-reported comparisons provide estimates for this short-term outcome. Their interpretation is limited by the small group sizes, the observed variability, and the uncertainty reflected in the confidence intervals.

The reported net mean differences are negative, but the registry does not state the subtraction convention. The confidence intervals for both comparisons include zero, and neither estimate should be treated as a precise measurement of the underlying difference. The p-values are from one-sided tests; the registry does not report the test direction rationale. These statistical summaries do not establish the clinical importance of the observed changes, and the record does not specify a threshold for a meaningful difference.

The endpoint is not a direct measure of long-term diabetes control, quality of life, complications, or later insulin requirements. It is a 24-hour insulin-dose change over a brief study interval. The record includes additional pharmacodynamic and pharmacokinetic outcomes, but their numerical results are not included in the registry record for this analysis, so no conclusions about those outcomes can be drawn from the primary endpoint summaries.

The safety counts are also limited in scope. No serious adverse events are reported in the groups, but the study’s small size and registry-reported observation do not establish the absence of risk. Other adverse events are counted, but their terms and clinical details are not reported in the registry record.

Overall, these results are best understood as early, exploratory evidence about measured insulin-dose changes during a controlled study period. They do not establish a durable effect, a definitive difference between doses, or a treatment recommendation. The registry record does not report enough information about missing data, multiplicity, or the prespecified statistical plan to resolve those additional interpretive questions.

15. Why This Trial Matters Statistically

This study illustrates how a continuous endpoint can be clinically recognizable but statistically uncertain. A mean change in insulin dose is easy to describe, yet its interpretation depends on the spread of individual changes, the comparator group, and the precision of the estimated difference. Reporting the means together with standard deviations and confidence intervals allows readers to distinguish the observed group averages from the uncertainty surrounding comparisons.

It also shows why the analysis population matters. The primary endpoint applies to participants who received study medication and had evaluable 24-hour insulin data. That definition is relevant to the result, but the registry record does not report the extent or handling of missing measurements. Without those details, it is difficult to assess whether the estimates represent all randomized participants or a subset with usable endpoint data.

The study’s small group sizes make caution especially important. A numerical separation between group means can coexist with broad uncertainty, and a p-value does not convey the range of effects compatible with the data. The reported confidence intervals include zero, making precision and direction central to interpretation rather than relying on a threshold-based reading of the p-values.

Finally, the registry lists several secondary outcomes in addition to the primary endpoint. That breadth makes the absence of a reported multiplicity strategy and analysis plan consequential for readers assessing the statistical evidence. The study is useful as an example of the distinction between early-stage descriptive evidence, sponsor-reported hypothesis tests, and a fully documented confirmatory analysis.

16. 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 trial was a small Phase 1 study, with 8.0 participants in each LY2409021 group and 4.0 in the placebo group for the primary endpoint. Such group sizes provide limited information about variability and make estimates potentially sensitive to individual observations. The broad confidence intervals reinforce that the size of the between-group differences is uncertain.

The primary outcome covers the change in 24-hour insulin dose from Day 1 to Day 2. It does not describe effects over longer follow-up. The registry summary describes a 7-day hospitalization period, but the primary endpoint itself is limited to the two stated study days. Longer-term outcomes and the persistence of any observed change are not reported in the registry record.

The registry does not report the randomization method, allocation concealment details, or who was masked. Although the study is recorded as double-masked, the masking procedures are not described. The reason for the unequal group sizes, the original sample-size justification, and the basis for the one-sided testing are also not reported in the registry record.

Missing-data methods and the number of participants without evaluable endpoint data are not reported. A multiplicity strategy for the primary and secondary outcomes is not reported either. The registry also does not report a statistical analysis plan, interim analyses, or a prespecified clinically important difference. These omissions prevent a full assessment of the relationship between planned and reported analyses.

The available primary summaries do not include participant-level data, adjusted estimates, or subgroup results. The sponsor-reported net differences lack an explicit subtraction convention in the record. Together, these limitations mean that the posted findings support description of the observed short-term results, but not a precise or comprehensive account of effects across other outcomes, populations, or time periods.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Study of LY2409021 in Participants With Type 1 Diabetes — NCT01640834.
• Registry record last updated 2018-10-29; status last verified 2018-03.
• Results first posted 2018-10-29.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

20. Record Summary

The registry record describes a completed, randomized, double-masked, parallel-group Phase 1 study of LY2409021 in participants with type 1 diabetes. It compared single oral doses of 100 mg and 300 mg with placebo. The primary endpoint was the mean change in 24-hour insulin dose from Day 1 to Day 2 among participants with evaluable data.

The posted group means were 1.85 insulin units for 100 mg, 0.8 for 300 mg, and 7.69 for placebo. Sponsor-reported net mean differences versus placebo were -5.84 and -6.89 insulin units, with 95% confidence intervals that included zero. The registry reports no serious adverse events in the groups, alongside other adverse event counts. Missing-data handling, multiplicity procedures, and the original sample-size justification are not reported in the registry record.