Anti-Inflammatory Small Drug Adjunctive Therapy for Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Anti-Inflammatory Small Drug Adjunctive Therapy for Type 2 Diabetes (NCT02597101), a randomized phase 2 study. The primary outcome is Insulin Sensitivity at 6 Months From Baseline.

2. Clinical Question

3. Trial Design

This was a phase 2 randomized, quadruple-masked, crossover trial. Its official title describes neutrophil elastase inhibition as adjunctive therapy in overweight and obese, insulin-resistant patients with type 2 diabetes. The intervention was AZD9668; the comparator was placebo. Both conditions included saxagliptin and optimized metformin as described in the registry.

The registry lists enrollment as 14. The primary outcome was insulin sensitivity at 6 months from baseline. The posted group summaries include 4 participants for placebo and 5 for AZD9668. The record does not explain how these reported outcome counts relate to enrollment, whether all randomized participants contributed data, or how the crossover structure was represented in the summaries.

A crossover design means participants receive more than one study condition at different times, so comparisons can use information from the same participants under different conditions. This is a general explanation of the design, not a claim about the exact sequence used in this trial. The treatment sequence, period lengths, washout procedures, and handling of period or carryover effects are not reported in the registry record.

The trial started in 2015-11, had primary completion on 2019-06-10, and is listed as completed. The record identifies 9 outcome measures and 1 posted analysis. It lists secondary outcomes involving adverse events, glycated HbA1c, oral glucose tolerance testing, an insulinogenic index using C-peptide, and serum inflammatory markers. Further detail about randomization procedures, masking roles, and the rationale for the planned enrollment is 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.

Before assignment: 14 subjects consented (9 screen failures, 5 randomized). Of the 5 randomized, 4 completed both arms while 1 was withdrawn after completing the AZD9668 arm and did not crossover to placebo.

Overall Study

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 Insulin Sensitivity at 6 Months From Baseline, measured over 6 months.

The primary endpoint was insulin sensitivity at 6 months from baseline, measured using the hyperinsulinemic-euglycemic clamp method. The registry defines the outcome as M/LBM: whole-body insulin sensitivity adjusted for lean body mass. It describes the calculation as steady-state glucose disposal in mL per kilogram of lean body mass divided by steady-state insulin concentrations in microIU/mL. The stated unit is (mL/kg/min)/(microIU/mL).

The clamp method evaluates glucose disposal under controlled insulin and glucose conditions. In general, a measure of insulin sensitivity is used to characterize how responsive the body is to insulin; it is distinct from a direct measure of a clinical event or from a measure of long-term disease complications. Adjusting glucose disposal for lean body mass aims to express the measure relative to that body compartment, while the insulin concentration is part of the reported calculation.

The outcome is therefore a continuous physiological measure, with higher or lower values interpreted in the context of its definition and the comparison being analyzed. The registry reports mean and standard deviation summaries for each condition. It does not report a confidence interval for the modeled comparison, and the posted information does not specify a clinically important threshold for interpreting a difference in this measure.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the group summary describe?

The primary measure is continuous, and the registry reports a mean and standard deviation for each condition. A mean describes the central value of the observed measurements. A standard deviation describes how dispersed measurements are around that mean; it is not a confidence interval and does not by itself describe uncertainty in the treatment comparison. The placebo mean was 9.9 with standard deviation 3.7 and 4 participants. The AZD9668 mean was 6.4 with standard deviation 3.5 and 5 participants.

These summaries are useful for describing the observed values, but they do not fully represent a crossover comparison. In that design, each participant may contribute measurements under different conditions. A comparison can therefore depend on paired observations and on how the model handles the design. The registry does not provide participant-level measurements or a detailed explanation of how the reported summaries were assembled.

What does a REML mixed model do?

The registry identifies the reported analysis as a REML mixed model. In general, a mixed model can account for both fixed effects, such as the treatment condition being compared, and random effects that represent variation among participants or other grouping structures. Restricted maximum likelihood, often abbreviated REML, is a method for estimating variance components in such models. These are general descriptions; the exact model specification for this trial is not reported in the registry record.

Mixed models can be appropriate for repeated or correlated measurements because measurements from the same participant are not necessarily independent. In a crossover design, a model may also need to represent relevant period, sequence, or carryover features. The registry does not describe which of these terms were included, what covariance structure was used, or how treatment order was handled. It is therefore not possible to reconstruct the model from the posted description alone.

How should the reported p-value be read?

The registry reports p <0.05 for the mixed model comparison. In general, a p-value measures how incompatible the observed data are with a specified null model, under the assumptions used for the analysis. It is not the probability that the null hypothesis is true, and it does not measure the size of an effect. The record does not provide a confidence interval for the modeled contrast, so the range of values compatible with the data is not available.

The listed model value is 0.05, but its statistical meaning is not labeled clearly in the posted analysis. It should not be treated as an interpretable effect estimate without knowing whether it represents a contrast, test statistic, or another model quantity. The record does not provide additional model output to resolve that ambiguity.

Why does precision matter?

A confidence interval is a range of estimates compatible with the data under a specified method. It conveys precision as well as direction and magnitude when reported for a contrast. The posted comparison has no confidence interval, and the group summaries come from small numbers of participants. Consequently, the inferential result gives a thresholded p-value but does not show how narrowly or broadly the underlying contrast is estimated.

In general, a small sample can yield imprecise estimates, and a model's results may depend on its assumptions. In this setting, the crossover structure makes within-participant data and period handling particularly relevant. Those details are not reported, so readers cannot evaluate whether the model captured all relevant dependence from the design.

What role does over-dispersion play?

Over-dispersion arises when observed variability is greater than a model's assumed variability. In general, failing to account for excess variation can make uncertainty appear too small and inferential results too strong. The primary endpoint is continuous, and the posted record does not report an over-dispersion assessment or indicate that over-dispersion was part of the analysis. The concept is relevant as a general modeling consideration, not as a claim that over-dispersion occurred in this trial.

What can planning calculations tell readers?

Sample-size planning calculations are illustrative rather than observed trial results. In general, they describe how many participants might be needed under specified assumptions about an effect, significance level, allocation, and model. They do not represent the power actually achieved by a completed trial, and a calculation based on an observed effect is not a substitute for a prespecified design justification. The registry record does not report a sample-size justification for the trial. Any planning illustration should therefore be interpreted as a hypothetical calculation, not as evidence about the trial's design rationale or its achieved power.

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("Placebo", "AZD9668")
n   <- c(4, 5)
m   <- c(9.9, 6.4)
sp  <- c(3.7, 3.5)
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  Placebo  4  9.900  3.700  4.012  15.788
CARM  AZD9668  5  6.400  3.500  2.054  10.746

11. Analyses Reported by the Sponsor

The sponsor's posted analysis is a REML mixed model comparison, with a reported model value of 0.05 and p <0.05. The group summaries are mean 9.9 with standard deviation 3.7 for placebo, based on 4 participants, and mean 6.4 with standard deviation 3.5 for AZD9668, based on 5 participants. The model result and the group summaries answer related but distinct questions: the summaries describe observed values by condition, while the model comparison is the reported inferential analysis.

No independent recalculated treatment contrast is available in the posted record. In particular, the summary statistics do not provide the participant-level crossover pairing needed to reconstruct a model-based comparison, and the model specification is not described in enough detail to reproduce it. The reported value of 0.05 is not labeled sufficiently to identify its meaning as an effect estimate. Since no discrepancy is identified, there is no conflicting recalculation to weigh against the sponsor's analysis. The sponsor's reported result should be described as reported, with its missing confidence interval and limited methodological detail kept in view.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists 9 outcome measures, including the primary insulin-sensitivity endpoint and secondary measures of adverse events, glycated HbA1c, oral glucose tolerance testing, an insulinogenic index using C-peptide, and serum inflammatory markers. It reports 1 analysis. The information available does not describe a multiplicity strategy, such as a prespecified hierarchy or adjustment across multiple outcome tests. The posted p-value should therefore be understood as the reported result for the stated comparison, rather than as evidence that the family of outcomes was controlled for multiple testing.

The primary outcome description identifies the measure and its 6-month timeframe, but does not specify the analysis population. The registry enrollment is 14, while the posted primary summaries include 4 participants for placebo and 5 for AZD9668. How those counts relate to participants randomized, participants who completed both conditions, or participants included in the model is not reported in the registry record.

Missing-data handling is also not reported in the registry record. The posted result does not describe whether any measurements were missing, how missing measurements were handled, or whether sensitivity analyses were conducted. In general, missingness can affect an estimate when the availability of measurements is related to participant characteristics or outcomes. That possibility cannot be evaluated without further information.

The record does not describe the analysis population, protocol deviations, or the exact participant flow into the model. It also does not report whether secondary outcomes were analyzed with separate models or how their results were selected for posting. These omissions limit interpretation of the inferential result and make it difficult to assess the completeness of the analysis across the registered outcomes.

13. 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.

For placebo, the registry lists 4 participants at risk for serious adverse events and 0 affected; it also lists 4 at risk for other adverse events and 0 affected. For AZD9668, it lists 5 participants at risk for serious adverse events and 0 affected; it also lists 5 at risk for other adverse events and 0 affected. Deaths are reported as 0 in both groups. No top serious adverse-event terms are listed.

These figures mean that the posted safety record contains no reported serious adverse events, other adverse events, or deaths in the listed groups. They do not establish that such events cannot occur, and they do not describe risks beyond the participants and observation period represented in the record. The duration covered by these safety counts and the criteria used to classify events are not reported in the registry record.

The numbers at risk are small, so the absence of reported events is limited information about safety. It does not provide a precise estimate of how often adverse events might occur in a broader population. These counts should be read as a description of the posted trial data, not as a general safety conclusion about AZD9668 or the background medicines.

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

The posted findings describe insulin-sensitivity measurements and a model-based comparison in a small phase 2 randomized crossover trial. The group mean was 9.9 for placebo and 6.4 for AZD9668, with standard deviations of 3.7 and 3.5, respectively. The sponsor's REML mixed model result is reported as a value of 0.05 with p <0.05. The record does not provide an interval estimate or enough model detail to identify the reported model value as a treatment-effect estimate.

These data do not, by themselves, answer whether the observed contrast is clinically meaningful. The registry does not state a threshold for a meaningful change in M/LBM, nor does the posted inferential result include a confidence interval that would show the precision of the estimate. The p-value conveys neither the magnitude of a treatment difference nor the probability that a treatment effect is real.

The findings concern the stated population, treatment conditions, and 6-month primary endpoint. They do not establish results for other populations, other treatment durations, or the listed secondary outcomes. The registry lists those secondary outcomes but the information summarized in the primary result does not report their findings.

The posted adverse-event counts report no affected participants in the listed categories and no deaths. Given the small numbers at risk and lack of further safety detail, these counts cannot establish absence of risk. More broadly, a completed trial and a statistically thresholded result do not resolve limitations arising from small outcome counts, uncertain analysis populations, missing model details, and the lack of a reported confidence interval. The appropriate reading is a limited description of the measurements and analysis recorded for this trial, not a definitive conclusion about treatment effects or safety.

15. Why This Trial Matters Statistically

This trial is statistically notable because its crossover design makes the relationship between repeated measurements central to interpretation. A crossover comparison can use each participant's experience under both conditions, potentially reducing the influence of stable differences between participants. That benefit depends on the actual sequence, timing, and analysis. The registry identifies the design but does not report the sequence, period structure, washout procedures, or model terms that would show how those features were handled.

The primary outcome is a continuous physiological measure assessed with a clamp method, rather than a clinical event. This makes clear reporting of the scale, calculation, and uncertainty especially important. The record defines M/LBM and gives group means and standard deviations, but it does not give an interval for the modeled contrast or identify the meaning of the model value of 0.05. Readers therefore have limited information about the size and precision of the comparison.

The enrollment is listed as 14, while the primary summaries include 4 participants for placebo and 5 for AZD9668. The record does not reconcile these figures or define the analysis population. In addition, the absence of a reported missing-data approach and multiplicity strategy leaves important questions about completeness and interpretation unanswered.

These are broadly useful statistical lessons: design labels alone do not establish how an analysis was carried out, and a p-value should be read alongside an effect estimate and its uncertainty. Here, the registry record supplies a concise account of the endpoint and a model result, but lacks details needed for a full independent assessment of the 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.

Several limitations are specific to the posted record for this trial. Although enrollment is listed as 14, the primary summaries include 4 participants for placebo and 5 for AZD9668. The record does not explain how these figures relate to enrollment, whether they represent participants with measurements in each condition, or how many contributed to the mixed model. This makes the analysis population difficult to identify.

The trial is described as a crossover study, but the record does not report treatment sequence, period duration, washout, or how period and carryover effects were handled. Those details matter when interpreting treatment comparisons in this design. The mixed model is named as REML, but the fixed effects, random effects, covariance structure, and other model specifications are not reported.

The result includes p <0.05 and a model value of 0.05, but no confidence interval or clearly identified contrast estimate. Thus, the inferential result does not show the precision or interpretable magnitude of the modeled comparison. The registry also does not report missing-data methods, sensitivity analyses, or a rationale for the planned enrollment.

The record lists multiple secondary outcomes but does not describe how multiplicity was addressed. It also does not provide detailed results for those outcomes in the information considered for the primary endpoint. Safety counts are reported, but the observation period and event-classification details are not reported in the registry record. These constraints do not negate the posted observations; they limit how fully the comparison and its broader implications can be assessed.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: Anti-Inflammatory Small Drug Adjunctive Therapy for Type 2 Diabetes — NCT02597101.
• Registry record last updated 2021-05-13; status last verified 2021-04.
• Results first posted 2021-05-13.
• 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

NCT02597101 is a completed phase 2 randomized, quadruple-masked crossover trial titled “Anti-Inflammatory Small Drug Adjunctive Therapy for Type 2 Diabetes.” It studied AZD9668 as an adjunct to saxagliptin and optimized metformin, compared with placebo alongside the same background medicines, in overweight and obese, insulin-resistant patients with type 2 diabetes.

The primary outcome was insulin sensitivity at 6 months from baseline, assessed by hyperinsulinemic-euglycemic clamp and expressed as M/LBM. The posted means were 9.9 for placebo and 6.4 for AZD9668. The reported REML mixed model comparison has a value of 0.05 and p <0.05, without a reported confidence interval. The registry lists enrollment of 14, while the primary summaries include 4 participants for placebo and 5 for AZD9668; the relationship between these counts is not explained. The record lists 9 outcome measures and 1 analysis.