Pre-Diagnosis GLP-1 Receptor Agonist Use and Post-Cancer Mortality in Adults With Obesity and Type 2 Diabetes: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Pre-Diagnosis GLP-1 Receptor Agonist Use and Post-Cancer Mortality in Adults With Obesity and Type 2 Diabetes (NCT07513259), a registered phase not stated study. The primary outcome is All-cause Mortality (Comparison 1).

2. Clinical Question

3. Trial Design

This is a completed observational study titled “Pre-Diagnosis GLP-1 Receptor Agonist Use and Post-Cancer Mortality in Adults With Obesity and Type 2 Diabetes: A Target Trial Emulation.” Its stated purpose is to examine whether diabetes treatment before cancer diagnosis is associated with survival and serious complications after obesity-associated cancer. The study uses de-identified electronic health record data from the TriNetX US Collaborative Network. Chung Shan Medical University is listed as the sponsor.

The registry describes a GLP-1 RA exposure cohort and a comparator cohort without GLP-1 RA prescriptions before cancer diagnosis, as well as treatment-specific comparisons. For the primary outcome, Comparison 1 is between GLP-1 RA use without prior SGLT2 inhibitor use and SGLT2 inhibitor treatment. Other listed comparisons involve DPP-4 inhibitors, sulfonylureas, metformin, and usual care. The study description says that comparison-specific propensity score-matched cohorts were used for the primary outcome. The matching variables, matching procedure, and balance results are not reported in the registry record.

The listed study start is 2018-01-01, the primary completion date is 2025-10-31, and the status is COMPLETED. Registry enrollment is 203424. That enrollment figure does not establish the number included in any particular matched comparison or endpoint analysis; comparison-specific denominators are not reported in the registry record. Allocation is not stated, and the registry explicitly characterizes the arms as not a like-for-like control comparison. As an observational analysis, the study evaluates associations in routine care and does not establish that treatment caused any outcome difference.

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: Enrollment and Participant Flow counts reflect the unique participants in the overall study population, classified into two mutually exclusive cohorts according to pre-diagnosis GLP-1 RA use. Participants were counted once in the Participant Flow module. Comparator-specific propensity score-matched cohorts were defined separately for prespecified pairwise outcome analyses and are reported in the relevant Baseline Characteristics and Outcome Measures sections.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Participants analyzed were members of the corresponding comparison-specific propensity score-matched cohort who were at risk at the start of the 36-month assessment window. Participants with the outcome before the assessment window were excluded. Therefore, the number analyzed may differ across outcomes.

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: Baseline characteristics are reported separately for each prespecified pairwise propensity score-matched comparison. For each baseline measure corresponding to a given comparison, data are provided only for the two arms included in that comparison; the remaining arms are assigned 0 participants analyzed in order to avoid counting participants more than once within a row.

Only the first 6 of 11 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as All-cause Mortality (Comparison 1), measured over From obesity-associated cancer diagnosis through 36 months.

The primary endpoint is all-cause mortality within 36 months after obesity-associated cancer diagnosis. It is recorded as the number of participants with the event, and its timeframe runs from cancer diagnosis through 36 months. The registry defines the primary outcome for the propensity score-matched cohort in Comparison 1, contrasting adults treated with GLP-1 RA without prior SGLT2 inhibitor use against adults treated with SGLT2 inhibitors.

All-cause mortality is a broad endpoint: it counts deaths without restricting the outcome to a particular cause. In general, this can capture survival regardless of whether a death is attributed to cancer, another illness, or another cause. It does not, by itself, explain why a death occurred or distinguish cancer-specific mortality from other mortality.

The registry states that participants analyzed were members of the corresponding matched cohort who were at risk at the start of the assessment window; those with the outcome before that window were excluded. It also notes that the number analyzed may differ across outcomes. However, the endpoint record does not provide comparison-specific denominators, event proportions, an effect estimate, confidence interval, or test result. The death counts listed by group elsewhere in the record are not accompanied by enough information to establish an endpoint risk or treatment contrast. Thus, the outcome definition is available, but its comparative result and precision are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

In general, a target trial emulation uses observational data to organize an analysis around a treatment question resembling one that might be posed in a trial. The label does not mean that participants were randomly assigned. Here, the registry identifies an observational study using electronic health records and describes treatment cohorts based on therapy before cancer diagnosis. It does not report a protocol specifying all features of a hypothetical trial, such as detailed eligibility rules, treatment strategies, or follow-up procedures. Those details are not reported in the registry record.

What does propensity score matching do?

In general, a propensity score represents the estimated chance of receiving a particular treatment given recorded characteristics. Matching attempts to pair or group treated and comparator participants with similar estimated treatment chances. The goal is to improve comparability on measured characteristics, not to guarantee that the groups are identical. Matching cannot balance characteristics that were not measured or not included, and it does not recreate random allocation.

The registry says that the primary outcome was assessed in a propensity score-matched cohort for Comparison 1. It does not report which characteristics entered the score, how matches were selected, whether matching was with or without replacement, how unmatched participants were handled, or whether balance was assessed. Without those details, readers cannot judge how well the method aligned the comparison groups. The record also does not describe the matching approach used, if any, for the other comparisons.

What does a count tell us?

The primary outcome is expressed as participants with an event, and the registry lists death counts for treatment groups. In general, a count describes the number of events observed, but it does not convey the frequency of the event without an appropriate denominator. Comparisons based on counts alone can be misleading when groups differ in size or in the number of people eligible for the assessment. The registry notes that people with an outcome before the assessment window were excluded and that the analyzed number may differ across outcomes. Comparison-specific denominators and mortality estimates are not reported in the registry record.

How should uncertainty be assessed?

In general, a confidence interval gives a range of values compatible with the data and analysis under its assumptions. A narrow interval indicates greater statistical precision than a wide one, although neither resolves bias in an observational design. A p-value, when reported, summarizes how compatible the data are with a specified null model; it is not a measure of clinical importance or proof of causation. No confidence interval or p-value for the primary endpoint is reported in the registry record, so the precision and statistical evidence for a contrast cannot be evaluated.

For an outcome measured over follow-up, analyses may use event counts, risks over a fixed period, or time-to-event methods. These approaches answer related but distinct questions and rely on different assumptions. The registry defines a 36-month endpoint window but does not state the statistical model or provide an outcome estimate. It is therefore not possible to determine from the record whether the reported death counts were analyzed as fixed-window risks, in a time-to-event model, or through another method.

What do the planning illustrations represent?

In general, sample-size calculations describe how much information a future study might need under stated assumptions. The record’s planning assumptions specify a two-sided significance level of 0.05, equal allocation, a normal approximation, and detection of a stated fraction of an observed effect. For time-to-event illustrations, the stated quantity is the total number of events; other illustrations use participants per arm. Such illustrations are planning exercises based on an observed effect, not observed power and not proof that the original study was designed to achieve a particular power. The registry does not report a sample-size justification from the study’s own design, and numerical planning estimates cannot be interpreted as enrollment or observed analysis results.

Overall, the available methods description supports a limited account: an observational electronic health record study, treatment-based comparisons, and propensity score matching for the primary comparison. Details needed to reproduce or fully assess the analysis—including the covariates, match diagnostics, outcome model, handling of incomplete records, and uncertainty estimates—are not reported in the registry record.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("GLP-1 RA Without History of SGLT2i (Comparison 1)", "SGLT2i (Comparison 1)")
n      <- c(12060, 12125)
events <- c(876, 1233)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  GLP-1 RA Without History of SGLT2i (Comparison 1)  12060  876  7.3  6.8  7.7  6.8  7.7  6.8  7.7
ARM  SGLT2i (Comparison 1)  12125  1233  10.2  9.6  10.7  9.6  10.7  9.6  10.7

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 25 outcome measures and 0 analyses. The outcome list includes all-cause mortality as the primary measure and secondary outcomes such as inpatient events and pericardial effusion across comparisons. The presence of multiple outcomes and treatment comparisons raises a general multiplicity issue: when many hypotheses are tested, the chance of at least one apparently notable result can increase. Methods such as prespecifying a primary hypothesis or adjusting for multiple testing can address aspects of this problem, but the registry does not report whether any such approach was used. No statistical analysis plan or multiplicity adjustment is reported in the registry record.

The record does not describe missing-data handling, the amount or pattern of missingness, or whether analyses used complete records, imputation, or another approach. Those omissions matter because incomplete electronic health records can affect which participants and outcomes are observed. The record also does not report exclusions beyond the endpoint description that participants with the outcome before the assessment window were excluded, nor does it give denominators for the analysis populations.

The primary endpoint population is described as the corresponding comparison-specific propensity score-matched cohort at risk at the start of the 36-month assessment window. The registry notes that the number analyzed can differ across outcomes. It does not report the precise analysis population for each outcome, the number excluded at each stage, or whether any sensitivity analyses were conducted. Thus, the extent to which missing information, exclusions, or multiple comparisons influenced the reported findings cannot be assessed from the registry record. These reporting gaps do not establish that a particular method was or was not used; they mean that it is not reported in the registry record.

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

The study addresses an important association question in a defined clinical context: treatment history before an obesity-associated cancer diagnosis and mortality afterward among adults with obesity and type 2 diabetes. Its 36-month all-cause mortality endpoint offers a clear time window, and the registry describes several treatment comparators in addition to the primary Comparison 1. This design can be useful for studying patterns of routine care, provided the analysis and its limitations are interpreted carefully.

The record does not report an estimated treatment contrast for the primary endpoint, its confidence interval, or its statistical test. The posted group death counts do not by themselves describe the probability of death in each cohort because denominators are not reported. Consequently, the record does not establish the direction, magnitude, or precision of an association. It also cannot show from these counts whether one therapy is associated with better or worse survival.

Even if an association estimate were available, an observational comparison would not alone demonstrate that the pre-diagnosis treatment caused a change in post-diagnosis mortality. Treatment groups may differ in recorded or unrecorded ways, and propensity score matching addresses only measured factors incorporated into its procedure. The registry record does not provide the details needed to judge the extent of confounding control.

The stated purpose includes survival and serious complications, and the listed outcomes include inpatient events and pericardial effusion. However, the record does not provide comparative estimates for these outcomes either. The appropriate conclusion is about what the record documents, not about treatment efficacy: it describes a completed observational analysis and its intended comparisons, but the available result information is insufficient to quantify the primary association or draw a causal conclusion.

13. Why This Trial Matters Statistically

This study matters statistically because it illustrates the distinction between an observational treatment comparison and a randomized trial. Electronic health record data can support questions about treatment patterns and later outcomes in routine care, but treatment groups are not necessarily exchangeable at the outset. Propensity score matching is one method for addressing measured differences; it does not remove the need to examine covariate selection, balance, missing information, and unmeasured confounding. The registry describes matching for the primary comparison but does not report the details needed to assess these issues.

It also highlights the importance of denominators and effect estimates. A list of event counts is not enough to compare risk when the analysis populations may differ. The registry explicitly notes that endpoint analysis populations can vary, while comparison-specific denominators and mortality estimates are not reported. Without them, neither the size nor precision of the association can be interpreted. Confidence intervals and clearly defined effect measures would be needed to describe uncertainty as well as direction and magnitude.

Finally, this record includes a broad outcome set and multiple treatment comparisons. That creates a need for transparent reporting of prespecified primary analyses and how multiplicity was handled. The record lists 25 outcome measures but does not report a multiplicity procedure. Together, these features make clear why a study’s question and design description should be distinguished from its evidentiary result. The topic is suitable for observational investigation, but interpretation depends on reporting enough information to evaluate the comparison and its uncertainty.

14. 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 principal limitation for interpreting the primary result is that the registry record does not report a comparative mortality estimate, confidence interval, p-value, or analysis denominator. The listed group death counts cannot be converted into risks or compared as rates without the corresponding analysis populations. The record also does not establish whether those counts refer exactly to the endpoint-specific 36-month analysis population.

The study is observational, and allocation is not stated. Its treatment groups may therefore differ in ways related to both therapy choice and subsequent outcomes. The registry says propensity score matching was used for the primary comparison, but does not identify the matching covariates, diagnostics, or handling of unmatched participants. Residual confounding cannot be assessed. The comparator descriptions and treatment-history details are also incomplete in the registry record, limiting evaluation of how the cohorts were defined.

Other important analysis details are not reported in the registry record: the statistical model, missing-data procedures, multiplicity approach, sensitivity analyses, and full analysis populations. The record lists 25 outcome measures and 0 analyses, leaving the relationship between listed outcomes and reported results unclear. Although it gives an overall enrollment of 203424, that number does not identify the number included in any particular matched comparison or endpoint analysis.

These limitations concern what can be evaluated from the registry record. They do not establish that a method was not used; they indicate that it is not reported in the registry record. The absence of comparative estimates prevents assessment of the direction, magnitude, and statistical precision of the primary association.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Pre-Diagnosis GLP-1 Receptor Agonist Use and Post-Cancer Mortality in Adults With Obesity and Type 2 Diabetes — NCT07513259.
• Registry record last updated 2026-08-17; status last verified 2026-06.
• Results first posted 2026-08-17.
• 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.

18. Record Summary

NCT07513259 is a completed observational target trial emulation sponsored by Chung Shan Medical University. It studies adults with obesity and type 2 diabetes who developed obesity-associated cancer, using de-identified electronic health record data from the TriNetX US Collaborative Network. Registry enrollment is 203424; the listed start date is 2018-01-01 and primary completion date is 2025-10-31.

The primary outcome is all-cause mortality from cancer diagnosis through 36 months. Comparison 1 contrasts GLP-1 RA use without prior SGLT2 inhibitor use with SGLT2 inhibitor treatment in a propensity score-matched cohort. Other listed comparisons involve DPP-4 inhibitors, sulfonylureas, metformin, and usual care. The record lists group death counts, but does not report endpoint denominators, comparative estimates, confidence intervals, or p-values. It therefore does not permit a quantitative conclusion about the association between pre-diagnosis treatment and post-cancer mortality.