Antihypertensive Medication Exposure as Risk for Impaired Glucose Tolerance: A PEAR Sub-Study: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Antihypertensive Medication Exposure as Risk for Impaired Glucose Tolerance: A PEAR Sub-Study (NCT01099397), a registered phase not stated study. The primary outcome is Number of Participants Diagnosed With Prediabetes or Normal Glucose, by 2 Measurements (Fasting Glucose Measurement and Glucose Measurement After a 2-hour Oral Glucose Tolerance Test [OGTT]), at Three Timepoints During Antihypertensive Treatment.

2. Clinical Question

3. Trial Design

This is a completed sub-study associated with PEAR, a parent study of antihypertensive treatment. The registry lists University of Florida as sponsor and 26 participants enrolled. The record gives a start date of 2009-05 and a primary completion date of 2010-12. Its stated purpose was to assess detection of prediabetes using a 2-hour OGTT compared with fasting glucose among a subset of patients treated with a beta-blocker and diuretic.

The design described is a single cohort, not a randomized trial comparing separate intervention and control groups. Participants had both fasting and 2-hour OGTT laboratory measurements. These methods were compared at baseline, 9 weeks, and 18 weeks after initiation of PEAR intervention(s). The registry describes the study population as participants eligible for PEAR, while the primary outcome population includes only participants with data at each time point. The number meeting that complete-data requirement is not reported in the registry record.

The registry lists the OGTT as an intervention of type “OTHER.” It does not describe a separate assigned treatment for this sub-study or state an allocation method. A like-for-like control comparison is therefore not the structure to use when reading the research question: the relevant contrast is between two diagnostic measurements in the cohort. Details of the parent study’s treatment allocation or procedures are not reported in the registry record and should not be inferred from the sub-study description.

The record marks the study completed and lists one primary outcome measure and three analyses. It does not report the outcome counts, analysis methods, or comparative estimates in the information available for this record. Completion status describes the study’s registry status; by itself, it does not show whether the measurement methods differed or how precisely any difference was estimated.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Only including participants with data at each time point

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 Number of Participants Diagnosed With Prediabetes or Normal Glucose, by 2 Measurements (Fasting Glucose Measurement and Glucose Measurement After a 2-hour Oral Glucose Tolerance Test [OGTT]), at Three Timepoints During Antihypertensive Treatment, measured over Baseline, 9 weeks, and 18 weeks after initiation of PEAR intervention(s).

The primary endpoint counts participants classified as having prediabetes or normal glucose using two measurements: fasting glucose and glucose after a 2-hour OGTT. The classifications were assessed at baseline, 9 weeks, and 18 weeks after initiation of PEAR intervention(s). The registry’s stated thresholds are fasting glucose above 99mg/dL and 2-hour OGTT glucose above 139mg/dL for prediabetes. The endpoint description says participants were followed through the parent study and had both laboratory measurements evaluated at each time point.

This endpoint addresses the study’s central measurement question: whether the 2-hour OGTT identifies prediabetes in a greater extent than fasting glucose in this subset of patients. Comparing both measures in one cohort makes the two classifications relevant to the same people and time points, rather than comparing unrelated groups. The endpoint is reported as numbers of participants by classification, so useful interpretation would require the counts for each method and time point, and ideally a direct comparison that accounts for the paired measurements.

The registry gives the endpoint definition and time frame but does not report its observed counts or comparative estimates. It also does not report how many participants had complete data at each time point. Consequently, the endpoint’s intended meaning can be described, but the posted record does not establish which method classified more participants or how consistent the classifications were.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation begins by separating the measurement question from a treatment-effect question. This sub-study compared fasting glucose and a 2-hour OGTT in the same cohort. It was not described as a randomized comparison between separate treatment groups. The question is whether the two methods classify participants differently, not whether assignment to one antihypertensive treatment caused an outcome relative to a control treatment.

Why does pairing matter?

When two methods are applied to the same participant, their results are linked. For each time point, a participant may be classified as prediabetic by both methods, normal by both, or differently by the two methods. The pattern of discordant classifications is especially relevant to a paired comparison. Aggregate totals for each method alone do not reveal how many participants changed category between methods. General statistical principle: methods for paired binary outcomes should preserve that within-person linkage; treating the totals as independent groups can misrepresent uncertainty.

The registry reports that participants had both measurements evaluated at three time points, but it does not report the cross-classification counts or the test or model used. Thus, it is not possible to assess agreement, quantify disagreement, or verify an inferential comparison from the record. Even if separate method-specific counts were available, they would not by themselves fully describe the paired data.

What does a confidence interval contribute?

General statistical explanation: a confidence interval expresses the range of effect values reasonably compatible with the observed data under the statistical model and its assumptions. It conveys precision as well as direction and magnitude. A narrow interval indicates greater precision than a wide interval, but neither width nor compatibility with a null value can be assessed without the estimate and interval. No confidence interval for the method comparison is reported in the registry record.

A p-value, if reported, would address how surprising data at least as inconsistent with a specified null hypothesis would be under the model. It would not measure the probability that the null hypothesis is true, nor would it show that a difference is important. The registry information does not report a p-value or describe the null hypothesis and test used. Statistical significance therefore cannot be assessed from this record.

How should repeated time points be considered?

Repeated measurements on the same people are also correlated over time. General statistical principle: an analysis of change or of an overall time pattern should account for the fact that observations from one participant are related. Separate counts at each time point can describe category frequencies, but they do not alone show each person’s trajectory or establish whether changes over time exceed measurement variation. The registry does not report a longitudinal model, a time comparison, or participant-level transitions.

The endpoint is binary in the sense that the registry describes counts of participants classified into glucose categories. The underlying laboratory measurements are converted into categories using the stated thresholds: fasting glucose above 99mg/dL or 2-hour OGTT glucose above 139mg/dL indicates prediabetes under this study definition. Categorization can make a result easier to interpret as a count, but it loses information about how far measurements lie from a threshold. The record does not report the underlying glucose distributions, so the effect of categorization on these results cannot be assessed.

What can the planning illustrations tell readers?

Planning illustrations are not observed outcomes and are not the trial’s own sample-size justification. The stated assumptions for the illustrations are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect expressed as a stated fraction of the observed effect. Those assumptions describe a planning exercise, not an analysis of the cohort’s paired measurement comparison. The registry does not report a sample-size justification for this sub-study, and planning calculations should not be read as observed power or evidence that the comparison was adequately powered.

In general, power calculations depend on the target effect, variability, design, and analysis assumptions. An illustration based on an observed effect is conditional on that effect estimate and is not a substitute for interpreting the observed estimate and its uncertainty. Since the comparative outcome estimate is not reported in the registry record, the illustrations cannot be independently interpreted as evidence about the measurement question. Their equal-allocation framing also does not describe the stated single-cohort, within-participant comparison.

Finally, a complete statistical assessment would need the analyzed counts, a method appropriate to paired binary classifications, an approach to repeated assessments if those are combined, and uncertainty estimates. The registry’s outcome definition and schedule are useful for understanding the intended analysis, but the actual analysis procedures and results are not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists three analyses, but their definitions, hypotheses, and results are not reported in the registry record. It is therefore unclear whether they correspond to separate time points, alternative comparisons, or other analyses. The record does not say whether the analyses were prespecified, whether one was primary among them, or whether any adjustment for multiple testing was planned or used. Without that information, the chance of false-positive findings across a set of analyses cannot be evaluated. This is a limitation in interpreting any isolated statistical test, were one to be reported.

The primary outcome population is described as including only participants with data at each time point. The number of participants meeting this condition is not reported in the registry record, nor are the counts missing at each assessment. The record does not describe reasons for missing measurements, whether participants with incomplete data were excluded from every analysis, or whether any missing-data method was used. It also does not report whether missingness was related to treatment, glucose status, or other participant characteristics.

General statistical principle: complete-case analyses can be biased when the availability of measurements is related to the outcome or factors associated with it. This possibility does not establish that bias occurred in this study; the registry record does not provide enough information to judge it. A comparison across time points can also have different analyzed populations if the complete-data rule is not applied consistently, but the analysis populations are not reported in sufficient detail.

The registry’s wording identifies an intended restriction to participants with data at each time point, but it does not establish how that rule was implemented for every analysis. Further details about exclusions, analysis populations, multiplicity handling, and sensitivity analyses are not reported in the registry record. These omissions limit assessment of both reproducibility and the robustness of any unreported comparison.

11. 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.

The registry records serious adverse event figures for the PEAR Sub-study Participants group. It lists 26 participants at risk and 0 participants affected by a serious adverse event. The separate “other” adverse-event fields also list 26 at risk and 0 affected. These are the recorded counts for the group, not comparative safety estimates between assigned treatments.

The study is described as a single cohort, and the registry does not identify a randomized control group for this sub-study. The serious adverse event figures therefore should not be read as a comparison of risks between interventions. The registry also does not provide event terms in the serious adverse event listing, so the kinds of events monitored or the detailed event definitions cannot be characterized from this record.

A recorded zero means no affected participants are listed in the registry for the stated field; it does not establish that risk is absent. With no event count above zero, a comparative estimate or precise event-rate assessment is not available from these entries. The record does not report additional safety analyses, follow-up detail specific to adverse events, or a comparison with another group. These figures should be interpreted as registry-reported counts, not as proof about safety beyond the scope of the recorded fields.

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

The study’s intended contribution is a direct comparison of fasting glucose and a 2-hour OGTT for classifying prediabetes in participants treated with a beta-blocker and diuretic. Both measurements were assessed in the same cohort at baseline, 9 weeks, and 18 weeks after initiation of PEAR intervention(s). The registry also specifies the thresholds used for the classification. These features define a clinically relevant measurement question in the population described by the record.

However, the registry record does not report the numbers classified as prediabetic or normal by either method, the paired classification patterns, or a comparative estimate. As a result, it cannot establish whether one method classified more participants, whether the methods agreed, or how precise any difference was. The study objective must not be mistaken for a reported result.

The record also does not establish that an apparent difference in classification, if observed, would represent a difference in underlying health rather than a consequence of using different measurements and thresholds. A diagnostic classification is not the same thing as a clinical outcome. The study description does not report subsequent clinical events, nor does it show that changing measurement practice would improve outcomes.

This is not evidence about the comparative effectiveness of antihypertensive treatments. The sub-study is not described as a randomized treatment comparison, and the outcome concerns glucose classification by two methods. The registry does not provide a basis for claims about treatment benefits or harms beyond the recorded adverse-event counts.

The most defensible reading is that the registry documents the design and intended endpoint of a small, single-cohort measurement comparison, while the results needed to answer its primary question are not reported in the registry record. Interpretation should remain limited to that distinction between the planned question and the available outcome evidence.

13. Why This Trial Matters Statistically

Clinical studies often rely on measurements that do not classify every participant in the same way. Comparing fasting glucose with a 2-hour OGTT in the same cohort is statistically useful because it makes the measurement methods the focus rather than comparing outcomes in unrelated groups. The PEAR sub-study was designed around that kind of question among participants treated with a beta-blocker and diuretic.

The registry’s description also makes clear why method, timing, and analysis population matter. Measurements were evaluated at baseline, 9 weeks, and 18 weeks, while the primary outcome population was restricted to participants with data at each time point. A sound interpretation would need to distinguish classifications at each assessment, account for the paired nature of the two measurements, and make clear which participants contributed to each analysis. Those details are not reported in the registry record.

The record is statistically informative about the intended comparison and its classification thresholds, but it does not report the outcome counts or estimates. That gap illustrates a central principle in evidence interpretation: a well-defined question and an appropriate-looking design do not substitute for reported results. Without estimates and their uncertainty, readers cannot judge the direction, magnitude, or precision of the comparison.

The study also highlights the difference between a planning calculation and an observed analysis. Planning illustrations rely on assumptions and are intended to explore design choices; they are not evidence that a particular result was observed or that the study had a particular level of power. They should not be used to fill in missing outcome results or to infer the effect of one measurement method relative to another.

For biostatisticians and clinical researchers, the record is a reminder to report paired classifications, analysis populations, missing-data handling, and uncertainty alongside endpoint definitions. Those elements are necessary to make a measurement comparison interpretable and reproducible.

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 interpretation is that the registry record does not report the primary outcome counts or the comparison between fasting glucose and the 2-hour OGTT. No estimate, uncertainty interval, or test result is available to assess whether classifications differed. The endpoint definition and time frame explain what was intended to be measured, but they cannot replace the observed results.

The complete-data population is not quantified. Although the primary outcome is described as limited to participants with data at each time point, the record does not state how many were included in that analysis or how missing measurements were handled. Without information on missingness and exclusions, the representativeness of the analyzed population cannot be assessed.

The registry does not report the paired classification counts needed to evaluate agreement or discordance, nor does it identify the statistical method used for the within-participant comparison. It also does not describe how repeated observations were incorporated into the analyses. The three listed analyses are not defined, and any multiplicity strategy is not reported in the registry record.

The record describes a single cohort and does not provide a randomized control comparison for the sub-study. The results, if available elsewhere, would need to be interpreted in the context of that design and the particular population of participants eligible for PEAR and treated with a beta-blocker and diuretic. The record does not provide enough detail to assess broader generalizability.

Finally, the registry’s planning assumptions are not a sample-size justification for the study’s actual design. The available information does not report a protocol rationale, statistical analysis plan, or interim analysis plan. These design and reporting details are not reported in the registry record, so they cannot be evaluated or inferred.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Antihypertensive Medication Exposure as Risk for Impaired Glucose Tolerance: A PEAR Sub-Study — NCT01099397.
• Registry record last updated 2018-04-03; status last verified 2018-03.
• Results first posted 2012-09-03.
• 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.

18. Record Summary

NCT01099397 is a completed PEAR sub-study sponsored by University of Florida, with 26 participants enrolled. The record lists a start date of 2009-05 and primary completion date of 2010-12. Its purpose was to compare prediabetes detection using fasting glucose and a 2-hour OGTT among a subset of patients treated with a beta-blocker and diuretic.

The primary outcome counted participants classified as prediabetic or normal by both methods at baseline, 9 weeks, and 18 weeks after initiation of PEAR intervention(s). The registry defines prediabetes as fasting glucose above 99mg/dL or 2-hour OGTT glucose above 139mg/dL. The outcome population was restricted to participants with data at each time point.

The comparative outcome counts, analysis methods, and uncertainty estimates are not reported in the registry record. The serious adverse event fields list 0 affected among 26 at risk. These entries describe the recorded counts and do not provide a comparative safety estimate.