This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Investigation Into the Effects of Blood Glucose Levels Upon Eating Behavior in Lean and Obese Non-diabetic and Diabetic Subjects (NCT02673203), a non-randomized na study. The primary outcome is Rate of Changing Glucose Level.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02673203 |
| Phase | NA |
| Condition | Diabetes |
| Allocation | Non-Randomized |
| Design model | Parallel |
| Masking | None (open label) |
| Enrollment | 31 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Yale University |
| Results posted | Yes |
2. Clinical Question
Population
The registry describes typical-weight healthy control participants and obese non-diabetic participants. The listed arm descriptions define typical weight as BMI <25 kg/m2 and obesity as BMI > 30 kg/m2. Although the title refers to diabetic subjects, the listed arms do not identify a diabetic group.
Intervention
Continuous Glucose Monitoring System (CGMS) measurements and a food log are listed. These are measurement approaches rather than treatments being compared for a clinical benefit.
Comparator
The registry lists typical-weight participants and obese non-diabetic participants. Allocation was non-randomized, so these groups should not be treated as equivalent groups created by random assignment.
Primary question
The primary endpoint is the rate of changing glucose level, measured in mg/dl/min. The broader stated question is whether daily glucose profiles relate to hunger levels and food intake.
The study’s stated aim was to investigate whether daily glucose profiles, assessed by continuous glucose monitoring for 1 week, relate to hunger and food intake in normal-weight and obese individuals. That is a question about association: whether glucose patterns and eating-related measurements vary together. It does not, by itself, establish that glucose patterns cause changes in hunger or food intake.
The registry’s primary outcome is narrower than that broader aim. It describes a rate of glucose change measured by a CGM sensor in interstitial tissue, with a stated timeframe of 3-5 days. The registry also names outcomes involving glucose peak and nadir, daily hunger rating, and nutritional intake. It does not report numerical results for these outcomes, so the direction, size, and precision of any association cannot be assessed from the registry record.
3. Trial Design
The registry describes a completed, non-randomized, parallel study sponsored by Yale University. Enrollment is listed as 31. The record gives a start date of 2016-02 and a primary completion date of 2017-12-14. The phase is listed as NA. These administrative details describe the study record but do not establish how participants were recruited, how long each participant took part, or whether all enrolled participants contributed data to every outcome.
The listed groups are Typical Weight Participants and Obese Non-diabetic Participants. Their descriptions define the former as BMI <25 kg/m2 and the latter as BMI > 30 kg/m2. The title refers to lean and obese non-diabetic and diabetic subjects, while the listed arm descriptions identify a healthy control group and an obese non-diabetic group. The registry record does not explain this difference or identify a diabetic arm in the listed group descriptions.
Allocation is NON RANDOMIZED and masking is NONE. As a result, the listed groups were not formed by random assignment, and the registry does not describe masking. Differences between the groups could therefore reflect other characteristics as well as glucose and eating-related measurements. The record does not describe covariate adjustment or other methods for addressing group differences.
The named measurement approaches are a Continuous Glucose Monitoring System (CGMS) and a food log. The summary describes monitoring daily glucose profiles for 1 week, while the primary outcome timeframe is listed as 3-5 days. The registry record does not clarify how these timeframe descriptions relate. It also does not specify the food-log procedures or the exact schedule for recording hunger and intake.
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
| Stage | Typical Weight Participants | Obese Non-diabetic Pariticpants |
|---|---|---|
| STARTED | 16 | 15 |
| COMPLETED | 16 | 15 |
| NOT COMPLETED | 0 | 0 |
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.
| Characteristic | Unit | Typical Weight Participants | Obese Non-diabetic Pariticpants | Total |
|---|---|---|---|---|
| Participants analysed | 16 | 15 | 31 | |
| Age, Continuous | years | 37.3 (SD 12.0) | 40.4 (SD 14.1) | 39 (SD 13) |
| Sex: Female, Male: Female | Participants | 13 | 12 | 25 |
| Sex: Female, Male: Male | Participants | 3 | 3 | 6 |
| Race/Ethnicity, Customized: Caucasian | Participants | 11 | 9 | 20 |
| Race/Ethnicity, Customized: African American | Participants | 1 | 2 | 3 |
| Race/Ethnicity, Customized: Hispanic | Participants | 2 | 3 | 5 |
| Race/Ethnicity, Customized: Asian | Participants | 2 | 1 | 3 |
| Region of Enrollment: United States | participants | 16 | 15 | 31 |
| Weight (kg) | kg | 67.8 (SD 12.0) | 90.6 (SD 12.6) | 79.3 (SD 21.8) |
| A1c (%) | % | 5.2 (SD 0.4) | 5.1 (SD 0.7) | 5.1 (SD 0.6) |
| BMI (kg/m2) | kg/m2 | 23.5 (SD 1.7) | 33.4 (SD 2.2) | 28.3 (SD 5.4) |
6. Primary Endpoint
The registry lists the primary outcome as Rate of Changing Glucose Level, measured over 3-5 days.
| Element | Registry entry |
|---|---|
| Result type | Least Squares Mean |
| Measure of spread | 95% Confidence Interval |
| Unit | mg/dl/min |
| Endpoint classification | Continuous Endpoint |
The primary endpoint is the rate of changing glucose level, expressed in mg/dl/min. The description says that the CGM sensor measures glucose levels from interstitial tissue and that the endpoint represents the rate of change in those measurements. The registry identifies the result type as a least-squares mean and its dispersion as a 95% confidence interval. It gives the measurement timeframe as 3-5 days.
A rate describes how quickly a measured quantity changes over time, rather than only its average level. In general, such a summary can help characterize glucose dynamics that a single measurement would not capture. It is relevant to the broader study question because the registry aims to relate daily glucose profiles to hunger and food intake. However, the record does not explain how individual rates were derived, what model was used, or how the reported least-squares mean was estimated.
The 95% confidence interval would ordinarily communicate uncertainty around the estimated mean. The numerical estimate and interval are not reported in the registry record, so neither the observed rate nor its precision can be described. The stated 3-5 days timeframe also differs from the summary’s description of monitoring for 1 week; the record does not explain that distinction.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry record identifies a primary endpoint and a completed study, but it does not provide the numerical primary result. No least-squares mean, 95% confidence interval, or inferential test result is reported. The record also lists 0 analyses. Therefore, the direction and magnitude of the glucose-rate outcome, and the precision with which it was estimated, cannot be evaluated from the registry record.
This absence of a reported estimate is not evidence that the groups had the same glucose-rate outcome. It is also not evidence that they differed. Without an estimate and its uncertainty, there is no basis for judging whether a difference is small or large, or whether the data are compatible with a range of effects. The same caution applies to the broader question about relationships between daily glucose profiles, hunger, and food intake.
The registry names secondary outcomes including glucose peak and nadir, average daily hunger rating, and energy, carbohydrate, fat, and protein intake. It does not report numerical results for these outcomes. Although the record lists 9 outcome measures, it lists 0 analyses. These are different kinds of information: an outcome-measure listing identifies what was intended to be measured or reported, while an analysis gives a statistical result. The outcome list alone cannot establish what was observed.
The study is non-randomized, so even if group-specific results were available, a difference would not automatically be attributable to group status or to glucose patterns. Group membership may be associated with other measured or unmeasured characteristics. The registry record does not report an adjustment strategy that would address those possibilities. The listed groups are useful context for understanding the study question, but they are not randomized treatment arms.
The planning assumptions describe sample-size illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation, and a normal approximation. Such calculations are planning illustrations, not observed power and not evidence of a result. Since the numerical effect and calculated sample sizes are not reported in the registry record, their practical implications cannot be assessed.
The safety listing reports zero serious events, zero other events, and zero deaths in each listed group. Those figures describe the recorded counts for the stated at-risk groups; they do not replace outcome results, and they do not resolve the unanswered question about glucose profiles and eating-related measures. Taken together, the registry describes the study’s aim and measurement plan, but it does not provide the statistical results needed to characterize what the study found.
9. Statistical Methods Explained
The statistical details available in the registry record are limited. It identifies the primary outcome as a least-squares mean with a 95% confidence interval and gives assumptions for sample-size illustrations. It does not report an analysis model, the numerical outcome estimate, confidence limits, test statistics, or analysis populations. The explanations below describe the general meaning of the methods named in the record; they are not claims about unreported steps in this study.
What does a least-squares mean represent?
In general, a least-squares mean is a model-based estimate of a mean. Unlike a simple arithmetic average, it is obtained from a statistical model and can reflect terms included in that model. For example, a model may account for specified participant characteristics or other design features. The registry record does not identify the model terms used here, so it is not possible to tell whether the primary estimate was adjusted for any characteristics, or what those characteristics might have been. The label alone does not establish that an adjustment was made.
The outcome is a rate of changing glucose level, in mg/dl/min. A rate expresses change relative to time; it is not the same as a glucose concentration at a single measurement. The endpoint description says the sensor measures glucose from interstitial tissue. The record does not explain the calculation used to turn sensor readings into a rate, how readings were handled, or whether a participant-level summary was calculated before the group estimate. Those details matter for interpreting exactly what the estimate represents.
How should a 95% confidence interval be read?
In general, a confidence interval describes uncertainty around an estimated quantity under the assumptions of the analysis. A narrower interval indicates greater precision than a wider interval, while an interval’s location indicates which values remain compatible with the estimate and its uncertainty. A confidence interval is not a guarantee that the true value lies within a particular interval, nor does it by itself establish clinical importance. Here, the registry names a 95% confidence interval but does not report its limits. Its width, location, and implications therefore cannot be evaluated.
A confidence interval and a significance test answer related but different questions. A test may evaluate a specified null hypothesis under a chosen model; an interval shows a range of uncertainty around an estimate. The registry record does not report a test or a null hypothesis for the primary endpoint. It would be inappropriate to infer a p-value or a conclusion from the confidence-interval label alone.
What does the non-randomized design mean statistically?
In general, random allocation helps make groups comparable on average with respect to factors that could affect an outcome. In a non-randomized study, group membership is not assigned by chance. The listed typical-weight and obese non-diabetic groups may therefore differ in characteristics beyond the category used to describe them. An observed group contrast, if reported, would be an association and could be influenced by such differences.
Statistical adjustment can sometimes address measured differences, but it cannot guarantee that all sources of confounding have been addressed. The registry does not describe an adjusted comparison, a covariate set, or a causal analysis. The primary endpoint is described as a least-squares mean, but that term alone does not show whether group comparisons were adjusted. The record also does not establish that the primary endpoint was intended as a between-group contrast rather than a separate estimate for a group.
What are the sample-size illustrations?
The registry describes planning illustrations computed from an observed effect, using a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is specified as a fraction of the observed effect. In general, these calculations estimate how many participants per arm might be needed under stated assumptions to detect a chosen effect with a target power. The assumptions are planning inputs; they are not a description of the enrolled sample’s observed statistical power.
For that reason, any sample-size figures should be read as planning illustrations, not as proof that the study was adequately or inadequately powered. The record does not report the underlying observed effect or the resulting sample-size values in the registry record for this analysis. It also does not report the trial’s own sample-size justification. Thus, the planning assumptions do not establish how the study was originally designed or what effect the study actually estimated.
What is not established by the available methods?
The record lists 9 outcome measures and 0 analyses. It does not describe the model used for the primary outcome, handling of repeated measurements, missing-data procedures, a multiplicity strategy, or analysis populations. These are not minor details: they can affect how estimates are formed, what uncertainty means, and how results across outcomes should be interpreted. Their absence from the registry record should be treated as missing methodological information, not as evidence that a particular method was or was not used.
In general, repeated measurements from the same participant may be statistically related, and an analysis may need to account for that dependence. Continuous glucose monitoring naturally concerns a profile of measurements over time, but the registry does not state how the readings were modeled or summarized. It is therefore not possible to say whether the analysis used a repeated-measures model, a participant-level summary, or another method. No further model assumptions can be attributed to the study from the information recorded.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists 9 outcome measures and 0 analyses. The named secondary outcomes include difference in glucose peak and nadir, average daily hunger rating, and nutritional intake for energy, carbohydrate, fat, and protein. The record does not report numerical findings for these outcomes or explain how the full set of measures relates to the primary question.
When several outcomes are examined, the chance of obtaining apparently noteworthy results can increase if each is tested independently. In general, a study may address this through a prespecified hierarchy, adjusted significance thresholds, or another multiplicity strategy. The registry record does not report whether any such approach was planned or used. There is also no reported testing plan from which to determine whether secondary results were confirmatory or exploratory.
Missing-data handling is not reported in the registry record. It does not say whether any glucose, hunger, or food-log measurements were missing, how missing observations were handled, or whether any sensitivity analysis was conducted. The enrollment figure alone cannot establish how many participants contributed data to a particular outcome.
Analysis populations are also not reported in the registry record. There is no description of inclusion rules for the primary or secondary analyses, or of how participants with incomplete measurements were treated. Consequently, it is not possible to determine whether analyses used all enrolled participants, a subset, or different populations for different outcomes. These reporting gaps limit assessment of the representativeness and robustness of any outcome estimates.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Typical Weight Participants | 0 | 16 | 0.0% |
| Obese Non-diabetic Pariticpants | 0 | 15 | 0.0% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Typical Weight Participants | 0 of 16 |
| Obese Non-diabetic Pariticpants | 0 of 15 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Typical Weight Participants | 0 of 16 |
| Obese Non-diabetic Pariticpants | 0 of 15 |
The registry lists safety counts separately for Typical Weight Participants and Obese Non-diabetic Participants. For the typical-weight group, the at-risk count is 16, with 0 serious events, 0 other events, and 0 deaths. For the obese non-diabetic group, the at-risk count is 15, with 0 serious events, 0 other events, and 0 deaths.
These figures mean that the registry reports no events in the listed categories among the stated at-risk participants. They should be read as recorded counts for these groups, not as proof that no symptoms or other experiences occurred. The registry record does not provide definitions of the event categories, details of how events were collected, or a description of the observation period for safety reporting.
The groups were not randomized, and the registry does not provide a comparative safety analysis. The counts therefore should not be interpreted as evidence of a difference or equivalence in safety between groups. They also do not answer the study’s primary question about glucose-rate measurements or the relationship between glucose profiles and eating-related outcomes.
12. What the Result Does, and Does Not, Mean
The registry record establishes that the study was completed and describes its broad aim: to investigate whether daily glucose profiles relate to hunger and food intake in normal-weight and obese individuals. It identifies CGM measurements, a food log, a primary glucose-rate endpoint, and several secondary outcome areas. It also describes the listed groups and their non-randomized allocation.
What the record does not establish is the direction or size of any observed relationship. The primary endpoint’s estimate and confidence interval are not reported, and the record lists 0 analyses. Numerical findings for hunger, food intake, glucose peaks, and nadirs are also not reported. Thus, the registry record does not show whether the measured outcomes were associated, how strong any association was, or how precise an estimate might have been.
Even if an association were reported, it would not by itself show that glucose changes caused hunger or eating behavior to change. In a non-randomized study, group differences and other characteristics could contribute to a pattern in the data. The registry record does not describe a causal analysis or an adjustment strategy.
The title refers to diabetic and non-diabetic subjects, but the listed arms identify a typical-weight healthy control group and an obese non-diabetic group. The record does not clarify this difference. It also describes glucose monitoring for 1 week in the summary and a 3-5 days timeframe for the primary outcome, without explaining how those descriptions fit together.
Accordingly, the record is informative about the intended question and measurement framework, but not about the study’s substantive statistical findings. The absence of reported results should not be read as evidence for or against a relationship between glucose profiles and eating-related measures.
13. Why This Trial Matters Statistically
This study is statistically interesting because it connects a continuously measured physiological profile with eating-related measures. A glucose rate of change describes a different feature of glucose data from a single concentration, and the stated question concerns how daily profiles relate to hunger and intake. That makes the endpoint and the timing of measurements important to interpretation.
The non-randomized parallel design is also central. Comparisons between typical-weight and obese non-diabetic participants can describe patterns across those listed groups, but group membership was not assigned by chance. An observed association would therefore need to be interpreted with attention to possible differences between groups, not as a treatment effect. The registry does not describe methods for addressing such differences.
The outcome list spans glucose dynamics, hunger, and several forms of nutritional intake. That breadth makes transparent reporting of the analysis plan, missing-data methods, and multiplicity strategy useful for judging how much weight to place on each finding. Those methods are not reported in the registry record. The record lists 9 outcome measures but 0 analyses, so it provides a measurement framework without the results needed to evaluate the statistical evidence.
Finally, the record illustrates why endpoint definitions and timeframes matter. The summary describes CGM monitoring for 1 week, while the primary outcome lists a 3-5 days timeframe. Without clarification, readers cannot determine how the primary endpoint’s observation window relates to the broader monitoring description. These limitations do not negate the study question, but they constrain the conclusions that can be drawn from the registry record.
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 most consequential limitation for interpreting this record is that it does not report the primary outcome estimate, its 95% confidence interval, or any outcome analyses. The registry lists 0 analyses. As a result, the statistical findings cannot be assessed for magnitude, direction, or precision.
The study is non-randomized, and the registry does not report an adjustment plan for differences between the listed groups. If group contrasts were estimated, unmeasured or measured differences could affect those contrasts. The record does not describe recruitment, eligibility details beyond the group BMI descriptions, or participant characteristics that would help assess comparability.
The title refers to diabetic and non-diabetic subjects, whereas the listed group descriptions identify typical-weight healthy controls and obese non-diabetic participants. The registry record does not clarify whether diabetic participants were included in another way or whether the title reflects a broader intended population. Generalizing the listed design beyond the groups described is therefore uncertain.
There is also an unresolved timeframe distinction: the summary refers to CGM monitoring for 1 week, while the primary endpoint lists 3-5 days. The record does not explain the sensor-reading schedule, how the glucose rate was calculated, how food logs were collected, or how hunger ratings were obtained. Missing-data and analysis-population procedures are not reported.
Finally, the registry does not describe a multiplicity strategy despite listing 9 outcome measures, and it does not provide the underlying effect or sample-size illustration values needed to assess the planning calculations. These omissions limit both methodological appraisal and interpretation of the intended analyses.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Investigation Into the Effects of Blood Glucose Levels Upon Eating Behavior in Lean and Obese Non-diabetic and Diabetic Subjects — NCT02673203.
- Registry record last updated 2022-05-05; status last verified 2022-04.
- Results first posted 2020-01-22.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT02673203 is a completed, non-randomized, parallel study sponsored by Yale University, with enrollment listed as 31. The record describes continuous glucose monitoring and food logging in typical-weight and obese non-diabetic groups. Its broader aim is to assess whether daily glucose profiles relate to hunger and food intake.
The primary endpoint is the rate of changing glucose level in mg/dl/min, reported as a least-squares mean with a 95% confidence interval and a timeframe of 3-5 days. The summary separately describes CGM monitoring for 1 week. The record lists 9 outcome measures and 0 analyses; no numerical outcome results are reported. Safety listings show 0 serious events, 0 other events, and 0 deaths in each of the listed groups.
The record therefore documents the study question, design, and measures, but does not provide the statistical results needed to judge the observed relationships or their precision.