The Glucosamine-study: The Effect of Glucosamine in Treatment of Chronic Low Back Pain: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for The Glucosamine-study: The Effect of Glucosamine in Treatment of Chronic Low Back Pain (NCT00404079), a randomized phase 4 study. The primary outcome is Roland Morris Disability Questionnaire.

2. Clinical Question

3. Trial Design

The Glucosamine-study is listed as a phase 4, randomized, quadruple-masked, parallel-group trial comparing glucosamine sulphate with placebo. The registry reports enrollment of 250 participants, with 125 in each arm. The record lists the study as completed. Its start is recorded as 2006-12 and primary completion as 2009-08.

The trial focused on chronic low back pain in patients with degenerative lumbar MRI findings. The registry’s brief description presents the rationale that lumbar osteoarthritis may have similarities to osteoarthritis in the knees and hips, and asks whether glucosamine sulphate might have symptomatic or structural effects in lumbar disease. That rationale is background for the study question, not evidence of an effect in this trial.

The intervention and comparator are listed as glucosamine sulphate and placebo. The arm descriptions contain no further treatment details. Dose, treatment schedule, duration of assigned treatment, adherence, and permitted co-interventions are not reported in the registry record. The record also does not specify which groups of people were masked, so the meaning of quadruple masking cannot be made more specific.

The primary outcome was the Norwegian version of the RMDQ, assessed at 1 year. Visual Analogue Scale and EuroQol-5D are listed as secondary outcomes. Results for those secondary outcomes are not included in the registry record for this analysis, so their estimated effects and precision cannot be described.

Randomization supports a comparison of assigned groups, while parallel allocation means participants remained in one assigned group rather than crossing between treatments. These general design features do not by themselves resolve missing-data or reporting concerns. Details of sequence generation, allocation concealment, recruitment flow, and the rationale for the planned enrollment are 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: Trial participation required no wash out, run-in or transtion phase. Patients were excluded if they fulfilled any of the exclusion criteria.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The number of participants for analysis followed the intention to treat principle. Imputation was performed with mulitple imputation.

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 Roland Morris Disability Questionnaire, measured over 1 year.

The primary endpoint was the score on the Norwegian version of the Roland Morris Disability Questionnaire at 1 year. The RMDQ is a self-administered, back-specific measure of pain-related disability. It has a 24-point scale, with higher scores indicating greater disability. The registry describes the measure as having content and construct validity and internal consistency, and as reproducible and sensitive to change over time for patients with low back pain.

The RMDQ is used because the clinical question concerns disability associated with back pain, not only pain intensity. A patient’s score expresses the level of reported functional limitation in a standardized format. The registry states that a reduction of 3 points in the total RMDQ score was classified in advance as a response to treatment. This is a response definition in the record; it is not the same as the between-group mean difference, and the posted arm summaries do not report how many participants met that response definition.

The posted primary-outcome summaries give a mean score of 9.0 and standard deviation of 4.0 for each arm, with 125 participants per arm. A mean summarizes the group’s average score, while the standard deviation describes variation among participant scores. These summaries describe scores at the stated 1-year timeframe; they do not, on their own, show individual change from baseline. Baseline scores and baseline-adjusted estimates are not reported in the registry record.

7. Results Posted in the Registry

• Glucosamine Sulphate: mean 9.00 versus 9.00 on control; difference 0.00 (95% CI -0.99 to 0.99), standardized effect d = 0.00; z-test p = 1.000.

8. Statistical Interpretation

9. Statistical Methods Explained

What does the unadjusted mean difference represent?

For a continuous outcome such as the RMDQ score, a mean difference compares the average score in one arm with the average score in the other. The recalculation uses the posted mean of 9.0 and standard deviation of 4.0 for each arm, with 125 participants per arm. Its direction is glucosamine sulphate minus placebo. A value of 0.0 means that the two posted averages are the same. It does not mean every participant had the same score, and it does not describe the distribution of individual changes.

This is an unadjusted comparison: it uses the posted arm-level summaries without adjustment for baseline measurements or other variables. The registry does not provide an adjusted RMDQ estimate with confidence limits in the material analyzed. The sponsor’s reported mixed models analysis is a different analysis, and its relationship to the arm-level continuous summaries is not made clear.

What does the confidence interval express?

General explanation: a confidence interval describes the precision of an estimate under the assumptions of the method used. A narrower interval indicates greater precision for the estimated contrast; a wider interval indicates more uncertainty. The recalculated 95% interval, from -0.992 to 0.992, includes zero and values on either side of zero. It is an interval for the unadjusted difference in means, not for the odds ratio listed in the sponsor analysis. It does not state the probability that a particular true effect lies inside the interval.

How should the p-value be read?

General explanation: a p-value measures how compatible data are with a specified null model, under the assumptions of the analysis. It is not a measure of clinical importance, and it is not the probability that the null hypothesis is true. The recalculated unadjusted p-value is 1.0 for the difference based on posted summary statistics. The sponsor’s analysis reports p = 0.05. Those values come from different reported analyses and should not be collapsed into one result. The registry does not provide enough methodological detail to reconcile them.

Why do randomization and masking matter?

General explanation: random assignment is intended to make treatment groups comparable on average at the start of a trial, including for factors that may not have been measured. Masking can reduce the influence of knowledge of assignment on participant reports, treatment delivery, or outcome assessment, depending on who is masked. This trial is listed as randomized and quadruple-masked, but the record does not identify the masked parties or describe sequence generation or allocation concealment. Those details are not reported in the registry record.

What is the role of the analysis population and imputation?

The registry says the number of participants for analysis followed the intention-to-treat principle and that multiple imputation was performed. General explanation: an intention-to-treat approach aims to analyze participants according to their randomized assignment, preserving the comparison created by randomization. Multiple imputation is a family of methods for handling missing outcome data by creating plausible completed datasets and combining results, while accounting for uncertainty in the imputed values. The registry does not report the amount or pattern of missing data, the imputation model, its assumptions, or sensitivity analyses. Consequently, the implementation and robustness of that approach cannot be assessed from this record.

What does over-dispersion mean for continuous outcomes?

General explanation: over-dispersion describes variability greater than a model assumes. It matters when the model’s uncertainty calculations depend on a particular variance structure, because underestimated variability can make intervals too narrow and tests too optimistic. The posted outcome is continuous and is summarized by means and standard deviations; no over-dispersion assessment is reported. The concept should not be taken as a finding about this trial.

How should the sponsor analysis be distinguished?

The registry lists a mixed models analysis with an odds ratio of 4.5 and p = 0.05, along with a separate entry noting a 95% confidence level but not giving confidence interval limits. It does not provide a clear endpoint label or contrast for those entries, and the group identifier is OG000. An odds ratio is not the same estimand as an unadjusted mean difference in RMDQ points. Without the model specification, outcome coding, contrast, or confidence limits, it is not possible to determine how the sponsor analysis relates to the primary-outcome summaries. Since the record flags a discrepancy, the sponsor’s own analysis takes precedence over the recalculation.

What can sample-size planning illustrations tell a reader?

General explanation: sample-size calculations explore how many participants might be needed under stated assumptions about an effect, variability, significance level, and allocation. The planning assumptions in the record specify a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the effect to detect defined as a stated fraction of the observed effect. Such illustrations are planning scenarios, not observed power and not evidence that the original trial was adequately powered. The trial’s own design rationale and sample-size justification are not reported in the registry record. The calculations should not be interpreted as the trial’s prespecified plan.

10. Confidence Intervals and Precision

The recalculated unadjusted mean difference in RMDQ score is 0.0, with a 95% confidence interval from -0.992 to 0.992. The interval is centered on the posted difference and spans values on both sides of zero. In practical terms, the summary-statistic calculation is compatible with a range of possible group differences rather than only an exact equality. The limits express uncertainty in the group mean contrast under the calculation’s assumptions.

The interval is expressed in RMDQ units. It concerns average scores, not the proportion of participants who improved, met the registry’s 3-point response definition, or experienced a particular score change. The record does not provide the number meeting the response definition. It also does not provide a baseline-adjusted estimate or participant-level change summaries from which those interpretations could be drawn.

The interval belongs to a simple unadjusted recalculation from the posted arm-level statistics. It is not the confidence interval for the mixed models result. Although the registry records a 95% confidence level for one sponsor-analysis entry, its lower and upper limits are absent. The sponsor analysis also reports an odds ratio, a quantity that cannot be read as RMDQ points. As a result, precision for the sponsor’s own estimate cannot be judged from the available interval information.

A confidence interval is not a guarantee about the true effect, nor does it demonstrate equivalence. Equivalence requires an appropriate margin and analysis, neither of which is reported in the registry record. The interval should therefore be read as the precision of the recalculated mean difference only, with the sponsor’s analysis taking precedence because the record flags a discrepancy.

11. Analyses Reported by the Sponsor

The sponsor’s posted analysis is listed as a mixed models analysis and reports an odds ratio of 4.5 with p = 0.05. A second entry reports a 95% confidence level but gives no lower or upper confidence limits. The listed group identifier is OG000. The registry does not explain which endpoint or contrast these entries represent, how the outcome was coded, or why an odds ratio is used alongside a primary outcome described as a continuous RMDQ score.

The unadjusted recalculation from the arm summaries is different in form: it compares mean RMDQ scores and gives a mean difference of 0.0, a 95% confidence interval from -0.992 to 0.992, and p = 1.0. These results are not a reconstruction of the mixed models analysis. A mean difference in scale units and an odds ratio are distinct quantities, and the record does not provide the model details needed to translate between them.

The registry flags a discrepancy. Accordingly, the sponsor’s own analysis takes precedence over the recalculation. That precedence does not make the sponsor result fully interpretable: the missing interval limits and unclear outcome contrast remain limitations of the record. The safest account is to state both analyses accurately and avoid treating the recalculated result as a correction or replacement.

12. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome, the RMDQ, and two secondary outcomes: Visual Analogue Scale and EuroQol-5D. It also reports 3 outcome measures and 2 analyses. Results for the secondary outcomes are not reported in the registry record, and the available analysis entries do not clearly map to the listed outcomes. The record does not describe a multiplicity adjustment, a hierarchy among outcomes, or a prespecified approach for interpreting multiple tests. Without those details, the effect of examining multiple outcomes or analyses on interpretation cannot be assessed.

The primary-outcome analysis population is described as following the intention-to-treat principle, with multiple imputation performed. General explanation: intention-to-treat analysis seeks to retain participants in their randomized groups, while imputation can address missing outcome values under specified assumptions. The registry does not report how many RMDQ values were missing, the reasons for missingness, or whether missingness differed between arms. It also does not give the imputation model, variables used, assumptions, or sensitivity analyses. These omissions prevent an assessment of how much the missing-data procedure may have influenced the estimate.

The record does not provide a participant flow accounting, the number included in each outcome analysis, or a comparison of participants with and without observed outcomes. Although the arm summaries each state 125 participants and the trial enrollment is 250, that information alone does not establish that all randomized participants had observed primary endpoint values; imputation is explicitly mentioned.

The registry’s sponsor analysis entries and the recalculated summary-statistic comparison are not clearly reconciled. Details of the prespecified statistical analysis plan, interim analyses, and sample-size justification are not reported in the registry record. No inference about these matters should be made from the reported values alone.

13. 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 reports 125 participants at risk for serious adverse events in each arm and 0 affected in each arm. It also reports 125 at risk for other adverse events in each arm and 0 affected in each arm. No serious adverse event terms are listed. These figures describe what is posted in the registry; they do not establish that no adverse events occurred outside the categories or reporting definitions used in the record.

Deaths are listed as null for both groups, rather than as a numeric count. Therefore, deaths are not reported in the registry record. A null entry should not be read as zero. The record does not provide further details about how adverse events were defined, the period over which they were collected, severity or causality assessments, or completeness of follow-up for safety.

These safety figures should be considered alongside the limited reporting detail. A count of zero affected participants in the posted categories is not by itself a comprehensive account of safety, and it does not establish that the treatments are safe or unsafe. It is a description of the registry entries only.

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

The posted arm summaries show equal RMDQ means of 9.0 and equal standard deviations of 4.0 at the 1-year timeframe, with 125 participants per arm. The recalculated unadjusted mean difference is 0.0, and its 95% confidence interval ranges from -0.992 to 0.992. These statistics describe the reported group averages and the precision of their simple comparison. They do not describe every participant’s experience or establish exact equivalence.

The RMDQ captures pain-related disability, with higher values indicating greater disability. The registry’s stated 3-point reduction is a response threshold, but the registry entries do not report how many participants reached it. A group mean comparison cannot substitute for that responder count. Nor can it show changes from baseline when baseline scores are not reported.

The registry separately reports a mixed models result as an odds ratio of 4.5 with p = 0.05, but the endpoint, contrast, and model details are unclear, and its confidence interval limits are missing. The discrepancy flag means the sponsor’s own analysis takes precedence over the recalculation. The two results should be kept distinct rather than described as if they were estimates of the same quantity.

The available record does not support conclusions about the secondary outcomes, subgroup effects, treatment adherence, or the robustness of imputation. It also does not supply details that would establish equivalence or resolve the discrepancy. The statistical account is therefore limited to describing the posted summaries and sponsor analysis, and the uncertainty and reporting gaps surrounding them. It is not medical advice and does not establish whether either treatment should be used.

15. Why This Trial Matters Statistically

This trial illustrates why a registry result should be read at the level of the estimand being reported. A mean difference in RMDQ points describes a contrast between average continuous scores. An odds ratio describes a different type of comparison. When a record presents both without a clear connection, the estimates cannot be treated as interchangeable or reconciled by looking only at their p-values.

It also shows the value of reporting confidence limits alongside estimates. The recalculated mean difference has a stated 95% confidence interval, which conveys uncertainty for that unadjusted contrast. The sponsor-analysis entry notes a 95% confidence level but omits the interval bounds, leaving its precision unclear. A p-value alone does not fill that gap.

The registry mentions intention-to-treat analysis and multiple imputation, but does not give the missing-data amount, imputation assumptions, or sensitivity analyses. Those details matter because readers need to understand how unavailable outcomes were handled and whether conclusions depend on assumptions about them. Similarly, listing secondary outcomes without their results or a multiplicity strategy leaves the overall pattern of evidence incomplete.

Finally, equal posted means do not prove equivalence. The response threshold in the registry, the group average contrast, and the sponsor’s mixed models analysis address different statistical questions or may do so; the record does not clarify their relationship. This trial is a useful example of why the outcome definition, analysis population, estimator, interval, and data completeness all need to be read together, while keeping unreported details explicitly unresolved.

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.

The registry record is incomplete for several trial details that would help evaluate the results. Eligibility criteria, participant characteristics, sequence generation, allocation concealment, and the identities of the masked parties are not reported in the registry record. Treatment dose, schedule, adherence, and arm-specific treatment descriptions are also not reported. These omissions limit assessment of how the trial was conducted and to whom its results may apply.

The primary endpoint is described at 1 year, but baseline RMDQ scores and baseline-adjusted results are not reported in the registry record. It is not possible to determine from the available summary whether the posted means reflect all randomized participants, observed values only, or an imputed analysis, despite the stated intention-to-treat approach and use of multiple imputation. Missing-data counts and methods are not reported in sufficient detail to assess this.

The sponsor’s mixed models analysis is difficult to interpret because the record gives an odds ratio of 4.5 and p = 0.05 without a clear endpoint contrast or model specification. A separate entry notes a 95% confidence level but omits interval limits. The discrepancy flag makes the sponsor analysis take precedence, but the information needed to understand its estimand and precision is not reported in the registry record.

The two secondary outcomes are named but their results are not reported in the registry record. The record does not describe multiplicity handling, interim analyses, or the trial’s sample-size justification. It also does not report the responder count for the stated 3-point RMDQ threshold. These are limitations of what can be concluded from the registry record, rather than findings that any particular method was or was not used.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: The Glucosamine-study: The Effect of Glucosamine in Treatment of Chronic Low Back Pain — NCT00404079.
• Registry record last updated 2011-04-29; status last verified 2011-02.
• Results first posted 2011-04-29.
• 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

The Glucosamine-study is a completed phase 4 randomized, quadruple-masked, parallel-group trial of glucosamine sulphate versus placebo in chronic low back pain with degenerative lumbar MRI findings. Enrollment was 250, with 125 participants in each arm. The primary outcome was the Norwegian RMDQ at 1 year; Visual Analogue Scale and EuroQol-5D were listed as secondary outcomes.

The posted RMDQ means were 9.0 in both arms, with standard deviations of 4.0. The recalculated unadjusted mean difference is 0.0, with a 95% confidence interval from -0.992 to 0.992 and p = 1.0. The record also reports a mixed models analysis with an odds ratio of 4.5 and p = 0.05, but its relationship to the primary outcome is unclear. Because the discrepancy flag is true, the sponsor’s own analysis takes precedence over the recalculation. Multiple imputation is reported, but its details and the secondary-outcome results are not reported in the registry record.