A Study of LY3437943 in Participants With Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of LY3437943 in Participants With Type 2 Diabetes (NCT04867785), a randomized phase 2 study. The primary outcome is Change From Baseline in Hemoglobin A1c (HbA1c).

2. Clinical Question

The clinical question is about the estimated change in a laboratory measure, not about a diagnosis or a clinical event. HbA1c reflects average plasma glucose over a prolonged period, according to the endpoint description. The registry reports adjusted least-squares means for change from baseline and sponsor analyses comparing treatment groups. These estimates describe the specified analysis population and follow-up period; they do not by themselves describe every enrolled participant or outcomes beyond the recorded timeframe.

3. Trial Design

The registry describes a randomized, double-masked, parallel-group Phase 2 study, with an enrollment of 281 and a completed status. The listed study start date is 2021-05-13 and the primary completion date is 2022-07-08. The brief summary says the study would last about 43 weeks. The primary endpoint timeframe is baseline to 24 weeks.

Participants were assigned to placebo, dulaglutide, or one of several LY3437943 dose regimens. LY3437943 was administered by once-weekly subcutaneous injection. Regimens included a 0.5 mg group and higher-dose groups with either a stated starting dose or stepwise escalation. The active comparator was 1.5 mg dulaglutide, administered once weekly by subcutaneous single-dose pen injection. Placebo was also administered once weekly by subcutaneous injection.

The available group summaries show that the numbers contributing to the primary endpoint differ by treatment group. The registry’s primary-outcome population definition further restricts the analysis to randomized participants with baseline and post-baseline measurements at the time point, excluding specified data after inadvertent enrollment, permanent study-drug discontinuation, or rescue-medication initiation. The record does not report the original allocation ratio, the detailed randomization procedure, or a full account of participant flow through follow-up.

The record identifies the study as Phase 2 and provides the trial’s interventions and outcome descriptions. It does not report the protocol’s original sample-size justification or the trial’s full statistical analysis plan. Those omissions matter when interpreting what comparisons were planned, how much emphasis was intended for each endpoint, and how the posted analyses handled multiple comparisons.

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: Four maintenance doses of LY3437943 were evaluated: 0.5 mg, 4 mg, 8 mg, and 12 mg. Dose escalation occurred in certain cohorts up to Week 12 during the 36-week treatment period. For maintenance doses equal to or greater than 4 mg, the initial dose will be 2 mg or 4 mg followed by additional escalation steps as appropriate.

Only the first 6 of 8 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized participants with baseline and post-baseline value at the specified time point excluding participants discontinuing study drug due to inadvertent enrollment and data after permanent discontinuation of study drug or initiation of rescue medication.

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

Only the first 6 of 9 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Hemoglobin A1c (HbA1c), measured over Baseline, 24 Weeks.

The primary endpoint is change from baseline in hemoglobin A1c (HbA1c), expressed as a percentage of HbA1c, at 24 weeks. The registry defines HbA1c as the glycosylated fraction of hemoglobin A and says it is measured primarily to identify average plasma glucose concentration over prolonged periods. Measuring change from baseline focuses the comparison on how the recorded value shifted during follow-up rather than only on the value at the end.

The posted arm estimates are least-squares means, with standard errors. The registry says they were calculated using a mixed model for repeated measures, with treatment, time, strata, and baseline terms in the model. In general, a least-squares mean is a model-based estimate that accounts for terms included in the specified model; it is not necessarily the same as a simple arithmetic average of individual changes.

The endpoint is a continuous measure, so the estimated treatment contrasts are expressed as differences in change. A more negative change indicates a greater decrease in HbA1c from baseline in the reported scale. Interpretation should remain tied to the analysis population and the 24-week timeframe. The registry’s outcome description does not make the endpoint a direct measure of long-term complications, nor does it establish what happens after the recorded follow-up.

7. Results Posted in the Registry

• 1.5 mg Dulaglutide: mean -1.41 versus -0.05 on control; difference -1.36 (95% CI -1.83 to -0.89), standardized effect d = -1.19; z-test p = <0.001.
• 0.5 mg LY3437943: mean -0.43 versus -0.05 on control; difference -0.38 (95% CI -0.95 to 0.19), standardized effect d = -0.27; z-test p = 0.190.
• 4 mg LY3437943 (2 mg): mean -1.39 versus -0.05 on control; difference -1.34 (95% CI -1.83 to -0.85), standardized effect d = -1.10; z-test p = <0.001.
• 4 mg LY3437943 (4 mg): mean -1.30 versus -0.05 on control; difference -1.25 (95% CI -1.85 to -0.65), standardized effect d = -0.96; z-test p = <0.001.

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 From Baseline in HbA1c

• 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:
  • 1.5 Milligram (mg) Dulaglutide: mean -1.41 versus -0.05 on control; difference -1.36 (95% CI -1.83 to -0.89), standardized effect d = -1.19; z-test p = <0.001.
  • 0.5 mg LY3437943: mean -0.43 versus -0.05 on control; difference -0.38 (95% CI -0.95 to 0.19), standardized effect d = -0.27; z-test p = 0.190.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for 1.5 Milligram (mg) Dulaglutide than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between 0.5 mg LY3437943 and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants Reaching HbA1c <7.0%

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • 1.5 Milligram (mg) Dulaglutide: 52.2% (24/46) versus 17.8% (8/45) on control; risk difference 34.4 percentage points (95% CI 16.1 to 52.6), risk ratio 2.93 (1.48–5.83); unadjusted two-proportion z-test p = <0.001.
  • 0.5 mg LY3437943: 32.6% (15/46) versus 17.8% (8/45) on control; risk difference 14.8 percentage points (95% CI -2.7 to 32.4), risk ratio 1.83 (0.86–3.90); unadjusted two-proportion z-test p = 0.104.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for 1.5 Milligram (mg) Dulaglutide than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between 0.5 mg LY3437943 and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participant Reaching HbA1c <7.0%

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • 1.5 Milligram (mg) Dulaglutide: 60.9% (28/46) versus 22.2% (10/45) on control; risk difference 38.6 percentage points (95% CI 20.0 to 57.3), risk ratio 2.74 (1.51–4.96); unadjusted two-proportion z-test p = <0.001.
  • 0.5 mg LY3437943: 37.0% (17/46) versus 22.2% (10/45) on control; risk difference 14.7 percentage points (95% CI -3.8 to 33.2), risk ratio 1.66 (0.86–3.23); unadjusted two-proportion z-test p = 0.124.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for 1.5 Milligram (mg) Dulaglutide than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between 0.5 mg LY3437943 and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary results broadly follow the primary HbA1c pattern: estimated changes and the proportion reaching the stated HbA1c threshold generally look more favorable in several LY3437943 groups than in placebo. The registry reports the threshold endpoint at Week 24 and Week 36. These outcomes are related to the primary measure, but they answer different questions: change in a continuous measure describes the estimated shift, while a threshold outcome describes the share meeting a specified cutoff.

Secondary outcomes can add context, but they should not be treated as independent confirmation of the primary result. In general, outcomes that are related to HbA1c may reflect overlapping information. Moreover, the registry lists fasting blood glucose and body weight among secondary outcomes, but corresponding results are not reported in the registry record in the result details available for this analysis.

The sponsor-reported secondary comparisons and the unadjusted recalculations are not always the same. As with the primary endpoint, sponsor-reported analyses take precedence where the registry flags a discrepancy. The registry does not report a multiplicity adjustment or a testing hierarchy for the secondary outcomes. When numerous endpoints, time points, and comparisons are considered, unadjusted results carry less weight because the chance of at least one apparently notable finding may increase across the collection of tests.

For that reason, secondary findings are best viewed as supportive descriptive evidence rather than definitive separate conclusions. Their interpretation also depends on the endpoint-specific analysis population, the amount of missing data, and whether analyses were planned in advance. Those details are not fully reported. The secondary results therefore complement the primary endpoint pattern but do not establish additional independent claims about all measured outcomes.

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

• Change From Baseline in HbA1c
• Change From Baseline in HbA1c
• Percentage of Participants Reaching HbA1c <7.0%
• Percentage of Participant Reaching HbA1c <7.0%
• Change From Baseline in Fasting Blood Glucose (FBG)
• Change From Baseline in Fasting Blood Glucose (FBG)
• Change From Baseline in Body Weight
• Change From Baseline in Body Weight

9. Statistical Interpretation

10. Statistical Methods Explained

What does the mixed model estimate?

The registry says the primary HbA1c outcome was analyzed with a mixed model for repeated measures. In general, this type of model uses repeated measurements over time and accounts for the fact that observations from the same participant are related. The stated model included treatment, time, strata, and baseline terms, with treatment-by-time, strata-by-time, and baseline-by-time components. The posted endpoint estimates are least-squares means, which are model-based estimates conditional on the terms included in the model.

This approach differs from simply subtracting two raw group averages. A model can use measurements at multiple post-baseline times and account for specified covariates and grouping factors. The record does not give every implementation detail, such as the covariance structure or the method used for handling all forms of missing measurements. Those details are not reported in the registry record, so the model cannot be fully reconstructed from the outcome description alone.

How should the treatment contrast be read?

A mean difference compares the estimated change in one group with the estimated change in the reference group. For the primary outcome, negative reported differences indicate a more negative estimated change in the LY3437943 group than in placebo. The sponsor’s reported LS mean differences and confidence intervals are the inferential results to prioritize because the record flags a discrepancy between those analyses and simple recalculations from posted summaries.

The recalculated contrasts are explicitly unadjusted and approximate. They use posted arm summaries rather than the sponsor’s model output. In general, a recalculation from group summaries cannot reproduce every feature of a repeated-measures analysis, its covariate adjustment, or its analysis population. Therefore, when the recalculated estimate, interval, or p-value differs from the sponsor’s reported analysis, the sponsor’s own analysis takes precedence.

What does a p-value say?

In general, a p-value describes how unusual data at least as incompatible with a null comparison would be if that null model and the analysis assumptions were correct. It does not measure the probability that the null hypothesis is true, the size of an effect, or its importance. The reported p-values should be read together with the estimates and confidence intervals. A small p-value does not remove uncertainty about the size of the estimated contrast.

The sponsor reported p = 0.188 for the 0.5 mg comparison and p < 0.001 for the other listed LY3437943 comparisons with placebo. The record does not say whether these p-values were adjusted for multiple comparisons. Given the several dose groups and other endpoints, that missing information limits how strongly the results can be interpreted as a family of tests.

Why do analysis populations matter?

The primary endpoint description defines a restricted analysis population: randomized participants with baseline and post-baseline values at the specified time, excluding participants discontinuing study drug due to inadvertent enrollment and excluding data after permanent discontinuation or rescue-medication initiation. In general, such restrictions can affect which observations contribute to an estimate and the question the estimate answers. An analysis that excludes post-discontinuation or post-rescue data does not necessarily estimate the same quantity as an analysis that follows all randomized participants regardless of subsequent events.

The registry does not provide enough information to assess the consequences of missing measurements, discontinuation, or rescue medication in detail. It does not report a complete accounting of missing outcomes for this endpoint or sensitivity analyses under alternative assumptions. Those aspects should not be inferred from the model label alone.

Why does multiplicity matter?

When many treatment comparisons or endpoints are examined, the chance of at least one apparently notable result can increase if each test is considered separately. In general, a multiplicity procedure can control the error rate across a specified family of tests. The record lists numerous analyses but does not report whether a formal adjustment, hierarchy, or other prespecified strategy was used. Consequently, the reported p-values should be treated as the values for the individual reported analyses, without assuming that they control error across all comparisons.

Overall, the model description supports a broad understanding of the primary estimate: it is a repeated-measures, model-based comparison of HbA1c change at the stated timeframe in a defined population. The registry does not provide the full specification needed to recreate or independently validate the sponsor’s model. That is why the posted sponsor results, rather than the simplified recalculation, are the appropriate basis for reading the primary analysis.

11. Confidence Intervals and Precision

The sponsor’s reported 95% confidence intervals communicate uncertainty around the estimated primary contrasts with placebo. In general, a confidence interval is a range of values compatible with the data and analysis under the method’s assumptions; it is not a guarantee that the true value lies within that particular interval. Wider intervals indicate less precision, while narrower intervals indicate greater precision, all else equal.

For 0.5 mg LY3437943, the reported difference was -0.38, with a 95% interval from -0.94 to 0.18. Because the interval includes zero, the data are compatible with values on either side of no difference under the reported analysis. For the other listed LY3437943 comparisons, the sponsor’s reported intervals were entirely below zero: -1.34 with interval -1.84 to -0.85; -1.25 with interval -1.85 to -0.65; -1.94 with interval -2.44 to -1.43; -1.83 with interval -2.42 to -1.24; and -1.96 with interval -2.43 to -1.50.

Those intervals should be read alongside the sizes of the contributing groups and the analysis definition. They quantify uncertainty under the sponsor’s analysis, not uncertainty from every possible source, such as model misspecification or missing-data assumptions. The record does not report sensitivity analyses that would show whether the conclusions change under alternative assumptions.

The recalculated intervals differ slightly from the sponsor-reported intervals. Because the registry indicates a discrepancy, the sponsor’s own analysis takes precedence. The recalculated intervals are approximate and unadjusted, and should not be used to override the model-based results. Neither intervals nor p-values address multiplicity unless the analysis explicitly accounts for it; the record does not report such an adjustment.

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 sample-size figures are planning illustrations based on observed effects, not the trial’s original design calculation. For these continuous and binary endpoints, the listed figures are participants per arm, not total participants across the study. The stated assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The planning scenarios use the stated fraction of the observed effect as the effect to detect.

The 80% power column gives the illustrative participants per arm for an 80% chance of detecting the specified effect under those assumptions, if the effect and model assumptions used for planning were correct. The 90% power column gives the corresponding number for a 90% chance. Higher planned power generally requires more participants for the same effect. The scenarios also show that smaller effects need more participants: planning for 75% or 50% of the observed effect yields larger figures than planning for the full observed effect.

For example, the primary-outcome planning entries vary across comparisons because the observed contrast and the estimated variability differ. The figures should therefore be understood as conditional illustrations, not a proposed design recommendation or a reconstruction of the study’s actual planning. The trial’s own original sample-size assumptions and justification are not reported in the registry record.

This is not observed power. It does not describe the probability that the completed study detected an effect, and it should not be used to explain a nonsignificant result after the fact. For time-to-event outcomes, the stated convention would be total events, but the displayed primary and secondary outcomes are not time-to-event outcomes.

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 = -1.360; pooled standard deviation σ = 1.147.
2. n = 2 × (1.96 + 0.842)² × 1.147² / -1.360² = 11.2, rounded up to 12 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 <- -1.36                      # observed difference in means
sigma <- 1.14718022642918                     # 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(4867785)
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 = 12 (80% power), 15 (90% power)
 75% of observed effect: n per arm = 20 (80% power), 27 (90% power)
 50% of observed effect: n per arm = 45 (80% power), 60 (90% power)
power.t.test: 12.2 
Simulated power at n = 12 per arm: 0.793 (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. Analyses Reported by the Sponsor

The registry contains sponsor-reported model analyses and simple recalculations based on the posted arm summaries. The sponsor used mixed models for the primary HbA1c comparisons; the recalculations are unadjusted approximations. The registry flags a discrepancy, so the sponsor’s own analysis takes precedence.

For example, the sponsor reports a difference of -0.38 for 0.5 mg LY3437943 versus placebo, a 95% confidence interval from -0.94 to 0.18, and p = 0.188. The recalculated contrast is -0.38, with an approximate interval from -0.948 to 0.188 and p = 0.19. The near agreement in the point estimate does not make the analyses interchangeable: their intervals and p-values are derived differently, and the sponsor’s model-based result is the primary inferential report.

For the other dose groups, the sponsor’s reported differences and intervals are also close to, but not identical with, the recalculated values. The discrepancy is particularly important because simple recalculations cannot reproduce all the features of a repeated-measures model or its defined analysis population. The same principle applies to secondary analyses: sponsor-reported analyses should be distinguished from unadjusted calculations made from arm-level summaries.

The registry does not explain the source of the discrepancy or provide enough detail to resolve it independently. Accordingly, recalculated values can help readers understand the direction and approximate scale of the contrast, but they should not replace the sponsor-reported analysis when the two differ.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists multiple treatment groups, comparisons, time points, and secondary outcomes. In general, testing many hypotheses without accounting for the full set can increase the chance of false-positive findings. A prespecified multiplicity strategy may define which comparisons form a family and how the error rate is controlled. The registry record does not report whether the primary or secondary analyses used an adjustment, a testing hierarchy, or another strategy for multiplicity.

This missing information limits interpretation of the collection of p-values. The sponsor’s reported p-values remain the analyses reported for their individual comparisons, but they should not automatically be read as error-controlled evidence across every dose and endpoint. This caveat is especially relevant for secondary outcomes, which are often numerous and may be exploratory; the registry does not state how these outcomes were positioned in the trial’s testing plan.

The primary outcome population is described as randomized participants with baseline and post-baseline data at the specified time, excluding participants who discontinued study drug due to inadvertent enrollment and data after permanent study-drug discontinuation or initiation of rescue medication. This definition informs what observations entered the reported analysis. The registry does not report enough detail to evaluate how many observations were excluded for each reason or how missing measurements were handled in all circumstances.

In general, missing-data methods rely on assumptions about why values are absent and what those values might have been. If participants with missing outcomes differ from those with observed outcomes, an analysis of observed data can be affected. The record does not report sensitivity analyses that test the robustness of the primary result to alternative missing-data assumptions, nor does it provide the full model specification needed to reproduce the analysis.

The record also does not report a complete analysis plan, details of interim analyses, or a full accounting of analysis populations for every endpoint. These are not grounds to presume that particular methods were or were not used. They are limits on what can be established from the registry record. Interpretation should therefore distinguish the reported model results from unreported multiplicity and missing-data procedures.

16. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: transurethral prostatectomy (4.5%), cardiac failure congestive (4.3%), cholecystitis chronic (4.2%), osteomyelitis acute (4.2%), basal cell carcinoma (4.2%).

The registry’s serious adverse event summaries give the number at risk and the number affected in each group. For placebo, the registry reports 3 affected among 45 at risk; for 1.5 mg dulaglutide, 1 among 46; for 0.5 mg LY3437943, 3 among 47; for 4 mg LY3437943 (2 mg), 1 among 23; for 4 mg LY3437943 (4 mg), 2 among 24; for 8 mg LY3437943 (2 mg), 2 among 26; for 8 mg LY3437943 (4 mg), 1 among 24; and for 12 mg LY3437943 (2 mg), 2 among 46. The registry records 0 deaths in each group.

These are group-level counts of participants with serious adverse events, not a complete description of every event, its timing, duration, or relationship to study treatment. The separate other-adverse-event counts are also reported by group, but they should not be combined with serious events as if they were the same category. The registry does not report event-level details or an adjusted comparison of safety between groups in the registry record.

Small group counts can be imprecise, and raw counts alone do not establish that an event was caused by a treatment. These summaries describe what the registry records during the study; they do not establish comparative safety beyond the reported follow-up.

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

The posted primary outcome indicates that the sponsor’s model estimated larger decreases in HbA1c relative to placebo for the listed LY3437943 dose groups other than 0.5 mg, whose reported confidence interval includes zero. This is a statistical description of the registry’s 24-week outcome analysis. It is not a recommendation about treatment, and it does not by itself establish the importance of a change for any individual participant.

The result does not show that every participant experienced the group-average change. Group estimates summarize data under the analysis model; individual outcomes can vary. Nor do the comparisons establish how LY3437943 dose groups compare directly with one another, because the primary sponsor analyses compare each listed dose with placebo. A pattern across estimates is not a substitute for a direct statistical test between doses.

The findings are also limited to the recorded endpoint timeframe and specified analysis population. They do not establish whether the estimated differences persisted after follow-up, how they relate to outcomes not measured in the registry results, or how the findings apply to people outside the study population. The record describes participants with type 2 diabetes who had inadequate control on diet and exercise alone or a stable dose of metformin; it does not support broader assumptions about other populations or treatment contexts.

Uncertainty remains about multiplicity, missing-data sensitivity, and some model details because these are not reported in the registry record. The sponsor’s own analysis takes precedence over the simple recalculations, but the registry does not provide enough detail to independently recreate that analysis. The results therefore support a bounded reading of the reported comparisons, not a complete account of every clinical or statistical question relevant to the interventions.

18. Why This Trial Matters Statistically

This study is statistically informative because it reports a model-based continuous primary endpoint across several active dose regimens, along with a placebo comparator and an active comparator. The different estimated changes across the LY3437943 groups make clear why readers should distinguish arm-level estimates from formal between-group tests. A visual or numerical pattern across groups does not itself prove a dose-response relationship.

The registry also makes a useful distinction possible between sponsor-reported analyses and simple recalculations from summary data. Their estimates are broadly similar in some comparisons but not identical. Recognizing that difference matters: model-based repeated-measures results use information and assumptions not recoverable from a simple calculation based on arm summaries. Where the record flags a discrepancy, the sponsor’s analysis takes precedence.

Precision is central to the interpretation. The sponsor-reported interval for the 0.5 mg comparison includes zero, while the intervals for the other listed dose comparisons are below zero. Such intervals show why effect estimates and uncertainty should be considered together rather than relying on p-values alone.

The trial also illustrates the limits of registry reporting. Multiplicity procedures, detailed missing-data sensitivity analyses, the full model specification, and original sample-size justification are not reported. A statistical reading can describe the posted results while keeping those omissions visible. This distinction helps prevent a set of individual comparisons from being treated as stronger or more comprehensive evidence than the record supports.

19. 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.

Several limitations are specific to the available record. The primary analysis uses a restricted population, excluding specified observations after permanent study-drug discontinuation or rescue medication and participants discontinuing because of inadvertent enrollment. The record does not show how many observations were excluded for each reason, so the effect of those restrictions cannot be assessed from the posted summaries alone.

The reported primary comparisons are model-based, but the registry gives only a concise model formula. Details needed for full replication, including the repeated-measures covariance structure and the handling of all forms of missing data, are not reported in the registry record. Sensitivity analyses addressing alternative assumptions are not reported either.

The registry does not report a multiplicity plan, so the individual p-values cannot be assumed to control error across the set of dose comparisons, outcomes, or time points. The original sample-size justification is also not reported; the planning illustrations based on observed effects are not a record of the trial’s original design rationale.

The primary endpoint is measured at 24 weeks, and the brief summary describes a study lasting about 43 weeks. The posted primary result therefore should not be interpreted as evidence about persistence beyond the measured endpoint. The data presented do not provide full results for every secondary outcome named in the registry; fasting blood glucose and body weight results are not reported in the registry record in the available outcome details.

Finally, group summaries and event counts cannot show the full distribution of individual outcomes or event-level safety context. These constraints do not invalidate the reported analyses, but they limit independent verification and the scope of conclusions that can be drawn.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of LY3437943 in Participants With Type 2 Diabetes — NCT04867785.
• Registry record last updated 2023-07-03; status last verified 2023-06.
• Results first posted 2023-07-03.
• 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.

23. Record Summary

The registry describes a completed, randomized, double-masked, parallel-group Phase 2 study of once-weekly LY3437943, placebo, and 1.5 mg dulaglutide in participants with type 2 diabetes. Enrollment was 281. The primary endpoint was change from baseline in HbA1c at 24 weeks, analyzed with a mixed model for repeated measures in a defined analysis population.

The sponsor reported differences from placebo for the LY3437943 dose groups. The reported interval for 0.5 mg includes zero; intervals for the other listed dose comparisons lie below zero. Because the record flags discrepancies between sponsor analyses and summary-statistic recalculations, the sponsor’s analyses take precedence. Multiplicity methods, detailed missing-data sensitivity analyses, and the original sample-size rationale are not reported in the registry record.