Effect of Calcium Glycerophosphate (CaGP) - Fluoride Dentifrice on Dental Biofilm pH: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effect of Calcium Glycerophosphate (CaGP) - Fluoride Dentifrice on Dental Biofilm pH (NCT00875212), a randomized phase 1 study. The primary outcome is Minimum pH.

2. Clinical Question

3. Trial Design

The registry describes a randomized, triple-masked, crossover study with four dentifrice conditions. Its brief summary calls the study double blind, whereas the design fields classify masking as triple. The registry record does not explain this difference. Twelve young adults took part in a 14-day, four-phase crossover study, with a 1-week wash-out period between phases. Each participant encountered the different dentifrice conditions in separate periods, so the design compares conditions within a participant rather than assigning one permanent treatment to each participant.

The four dentifrices were a control with no fluoride and no calcium glycerophosphate, calcium glycerophosphate only at 0.13%, fluoride only at 1500 ppm, and calcium glycerophosphate plus fluoride at those stated concentrations. The brief summary says coded dentifrices were randomly assigned to volunteers. The arm description confirms use of four dentifrices in four periods, but further details of the sequence generation, allocation concealment, adherence, and wash-out assessment are not reported in the registry record.

Biofilm pH was measured at a single site using a microelectrode after a cariogenic challenge with a 10% w/v sucrose solution. The summary describes readings at baseline and at 1, 7, 14, and 21 minutes after the challenge. It also describes measurement sets before dentifrice use, 1 minute after use, and after 7 days and 14 days of use at 12 hours after brushing. Stephen curves and mean area under the curve were obtained, although the posted primary outcome is minimum pH. The registry lists the study as Phase 1 and completed. The start date is 2006-01 and primary completion is 2007-12.

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: As a pre-assignment procedure, all volunteers signed an informed consent. The the volunteers received a dental hygiene kit containing a toothbrush and a fluoride-free and calcium-free dentifrice as part of the 1-week wash-out period prior to group assignment.

Placebo Dentifrice

CaGP Only Dentifrice

Fluoride Only Dentifrice

CaGP and Fluoride Dentifrice

Analysis population for the primary outcome, as defined in the registry: The number of subjects were determined by a pilot study carried out in 3 volunteers. The study has a crossover design. Thus the 4 groups of dentifrices tested included the same subjects in a different time-measurement avoiding the influence of individual variables in the analysis.

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 Minimum pH, measured over at 1 minute (minimum fermenting pH) or at 7 minutes.

The primary endpoint is titled “Minimum pH.” It concerns the lowest pH observed in dental biofilm after participants received a sucrose challenge. The outcome description says pH was measured in vivo with the microtouch method using a palladium microelectrode and a reference electrode. The brief summary also describes a single-site microelectrode measurement and readings at multiple times after the challenge.

Minimum pH is used to characterize the low point of the biofilm pH response after exposure to sucrose. As a general statistical and measurement concept, a minimum captures an extreme point in a sequence of measurements, rather than the average pH across the entire sequence. It can therefore answer a different question from an area-under-the-curve summary, which reflects the pattern over time. The registry says that Stephen curves and mean area under the curve were obtained, but the posted primary outcome is the mean of the lowest pH observed at each time after sucrose use.

The stated timeframe is “at 1 minute (minimum fermenting pH) or at 7 minutes.” That wording leaves uncertainty about whether the reported minimum is specifically the 1-minute value, the 7-minute value, or a measure that involves both. This endpoint timing is not resolved in the registry record. The outcome is a continuous measure, reported as a mean with standard deviation. Its value describes biofilm pH under the study conditions; it is not, by itself, a measure of dental caries incidence or a clinical outcome over time.

7. Results Posted in the Registry

• Calcium Glycerophosphate: mean 5.48 versus 5.40 on control; difference 0.08 (95% CI -0.42 to 0.58), standardized effect d = 0.13; z-test p = 0.754.
• Fluoride: mean 5.25 versus 5.40 on control; difference -0.15 (95% CI -0.70 to 0.40), standardized effect d = -0.22; z-test p = 0.595.
• Calcium Glycerophosphate and Fluoride: mean 5.70 versus 5.40 on control; difference 0.30 (95% CI -0.12 to 0.72), standardized effect d = 0.57; z-test p = 0.159.

8. Statistical Interpretation

9. Statistical Methods Explained

What does a mean difference describe?

A mean difference expresses the difference between the average endpoint value in one condition and the average value in a reference condition. In the recalculations, the control mean is compared separately with each of the three other dentifrice means. A positive estimate indicates a higher mean pH in the named dentifrice condition than in control; a negative estimate indicates a lower mean. The estimates are 0.08 for calcium glycerophosphate, -0.15 for fluoride, and 0.3 for the combination. These are unadjusted recalculations based on posted arm summaries, not the sponsor’s stated analysis.

The arm summaries also give a standard deviation for each condition. In general, a standard deviation describes the spread of observed values around a mean. It is not the same as the uncertainty in the estimated mean or mean difference. The recalculations use the arm means and standard deviations to estimate pairwise differences and their intervals. Such calculations do not recreate an analysis that accounts for the paired or repeated structure of a crossover trial unless that structure is explicitly included. The registry record does not report enough detail to establish whether these recalculations reproduce the original analysis model.

What does the sponsor’s ANOVA tell the reader?

ANOVA is a general method for assessing whether there is evidence of differences across multiple group means or conditions, according to the model used. The registry reports an ANOVA across the four dentifrice groups, labels the parameter “Median Difference (Final Values),” gives a value of 5.5, and reports p = 0.001. It also states a 95% confidence level but supplies no interval limits. The wording combines an ANOVA method with a parameter label referring to median difference, and the record does not explain how these terms relate to one another. The same result is listed twice. The registry record does not provide the model specification, the exact test statistic, degrees of freedom, or adjusted estimates.

A group-level analysis and pairwise comparisons answer related but different questions. A group-level test concerns evidence of a difference somewhere among the conditions under its model. A pairwise comparison estimates the difference for one specified pair. A result from one form of analysis does not automatically identify which pair differs, and pairwise estimates with broad intervals can coexist with a group-level result from a different model or statistic. Because the discrepancy flag is true, the sponsor’s own analysis takes precedence over the simple recalculations when describing the reported trial analysis.

Why does the crossover structure matter?

In a crossover study, participants receive different conditions in different periods. General statistical reasoning indicates that this structure can account for stable participant-to-participant differences by comparing measurements within the same participant. It also means that observations from one participant are related, rather than independent in the same way as observations from separate groups. Period effects, sequence effects, carryover, and the adequacy of wash-out can be relevant considerations in crossover analysis. The registry describes random assignment, four phases, and a 1-week wash-out between phases, but does not report a statistical model for these design features or results of checks for carryover or period effects.

The simple recalculations use the posted arm summaries and are unadjusted. They should therefore be understood as summary-statistic comparisons, not as a reconstruction of a crossover model. The registry record does not report paired participant-level differences, their variability, or the correlation between measurements across periods. Without those details, it is not possible to tell from the posted record whether a participant-paired analysis would produce the same estimates or precision.

What does a p-value mean in this setting?

In general, a p-value describes how incompatible observed data are with a specified null hypothesis under the assumptions of a statistical model. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The recalculated p-values are unadjusted pairwise values: 0.754 for calcium glycerophosphate, 0.595 for fluoride, and 0.159 for the combination. The sponsor’s reported p-value is 0.001 for the listed ANOVA result. These values belong to different analyses, so they should not be compared as if they were generated by one common model and one common question.

Multiplicity refers generally to the increased chance of false-positive findings when many hypotheses are tested. The study has four conditions and reports more than one outcome measure, while the registry posts a group-level analysis and the recalculations include three pairwise comparisons. The registry record does not report a multiplicity adjustment or a prespecified testing hierarchy. Consequently, the status of the reported p-values with respect to multiple testing is not reported in the registry record.

What information is absent?

To assess the analysis fully, a reader would need a clear definition of the analysis population, the endpoint time point, the model used for the crossover design, missing-data handling, and the plan for multiple comparisons. The registry record does not report those details. It also does not report an explanation for the duplicate ANOVA entries or confidence-interval limits for the sponsor’s analysis. These omissions limit how precisely the reported result can be interpreted, even though the registry does provide descriptive means, standard deviations, and the stated p-values.

10. Confidence Intervals and Precision

The recalculated intervals convey uncertainty around the unadjusted mean differences. For calcium glycerophosphate versus control, the estimate is 0.08 and the interval extends from -0.420 to 0.581. For fluoride versus control, the estimate is -0.15 and the interval extends from -0.702 to 0.402. For the combination versus control, the estimate is 0.3 and the interval extends from -0.118 to 0.718. Each interval spans zero and includes values in both directions. The ranges are therefore compatible with a range of possible mean differences, including no difference, under the recalculation assumptions.

In general, a confidence interval communicates both the estimated effect and its precision: narrower intervals indicate greater precision, while wider intervals leave more uncertainty about the size and direction of an effect. These intervals are not evidence that every value in the range is equally likely, and they do not directly express clinical importance. The registry does not state the confidence level for these recalculated pairwise intervals, so no confidence level should be assigned to them from the posted record alone.

The sponsor’s analysis states a 95% confidence level but gives no lower or upper interval limits. Its reported value is 5.5, with p = 0.001. Without the interval limits and clearer specification of the statistic, the precision of that result cannot be assessed from the registry record. The sponsor result is also a group-level ANOVA entry, whereas the recalculated intervals are pairwise mean-difference intervals. Their precision cannot be compared directly as if they represented the same estimand.

Precision is also affected by the design and by the information available for analysis. The record gives 12 participants and repeated conditions in a crossover design, but it does not provide participant-level paired variability, missingness, or the crossover model. The simple intervals therefore should be read as unadjusted recalculations from posted summaries rather than as the definitive uncertainty estimates for the trial’s original analysis.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are illustrations based on the observed effects, not the original trial’s design calculations. They assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. For each comparison, the assumed effect to detect is a stated fraction of the observed effect: 100%, 75%, or 50%. The trial’s original sample-size assumptions are not reported in the registry record. These planning figures are participants per arm, not total event counts.

The 80% power column gives the planned number of participants per arm for a hypothetical study to have an 80% chance of detecting the specified effect under those assumptions, if the effect and model assumptions were correct. The 90% power column gives the corresponding number for a 90% chance. These are design targets for future planning, not probabilities that the observed trial result is correct. The listed figures vary by comparison because the observed effects and variability differ. For example, planning for 75% of the observed effect gives 1,707 participants per arm for calcium glycerophosphate, 591 for fluoride, and 85 for the combination at the 80% power target. The corresponding 90% power figures are 2,284, 792, and 114 participants per arm.

Smaller effects are harder to distinguish from random variation and generally require more participants under otherwise similar assumptions. The planning figures reflect that principle: the entries for 50% of the observed effect are larger than those for 100% of the observed effect for each comparison. The estimates should be treated as illustrations rather than prescriptions because the calculation uses simple two-arm planning assumptions and the observed effect from a small crossover study. It does not incorporate an original protocol’s design assumptions, and those assumptions are not reported in the registry record. This is not observed power.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 0.080; pooled standard deviation σ = 0.625.
2. n = 2 × (1.96 + 0.842)² × 0.625² / 0.080² = 959.6, rounded up to 960 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 0.0800000000000001                      # observed difference in means
sigma <- 0.62549980015984                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(875212)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 960 (80% power), 1,285 (90% power)
 75% of observed effect: n per arm = 1,707 (80% power), 2,284 (90% power)
 50% of observed effect: n per arm = 3,839 (80% power), 5,139 (90% power)
power.t.test: 960.6 
Simulated power at n = 960 per arm: 0.798 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Analyses Reported by the Sponsor

The sponsor’s posted analysis is an ANOVA across all four dentifrice groups. It reports a value of 5.5, labels the parameter “Median Difference (Final Values),” states a 95% confidence level, and gives p = 0.001. No confidence-interval limits are reported, and the same result appears twice. The registry record does not explain the duplicate entry or clarify the meaning of the statistic label alongside the ANOVA method.

The recalculated results are different in scope: they are three unadjusted pairwise mean differences comparing each active dentifrice condition with control. Their estimates are 0.08, -0.15, and 0.3, with intervals that include zero. These calculations use the posted means and standard deviations, and do not reproduce a reported crossover model. A group-level ANOVA result and separate pairwise mean-difference recalculations are not interchangeable analyses; they may target different summaries and use different assumptions.

Because the discrepancy flag is true, the sponsor’s own analysis takes precedence as the trial’s reported analysis. The recalculations can help describe the arm summaries and illustrate pairwise uncertainty, but they should not replace the sponsor’s result or be used to claim that it is disproved. The registry record does not include enough methodological detail to reconcile the findings, specify the ANOVA model, or determine how the crossover and multiple comparisons were handled.

14. Multiplicity, Missing Data and Analysis Populations

The registry record does not report a multiplicity plan. This matters because the trial includes four dentifrice conditions and reports an ANOVA result, while the unadjusted recalculations present three comparisons against control. In general, testing several contrasts creates more opportunities for at least one result to appear unusual by chance. Methods such as a prespecified testing hierarchy or an adjustment can address that issue, but the registry record does not say whether any such approach was planned or used. The recalculated p-values are explicitly unadjusted.

The record also does not report a clear analysis population. It lists 12 participants, but does not state whether all participants contributed to every condition and measurement, how protocol deviations were treated, or whether results were based on an intention-to-treat or per-protocol population. Missing-data methods and the amount or pattern of missing data are not reported in the registry record. Therefore, the influence of missing observations on the reported estimates cannot be assessed.

For a crossover design, the analytic method should account for the repeated condition measurements within participants and may need to consider period and sequence effects. The registry describes the crossover structure and wash-out period but does not report whether the ANOVA model accounted for those features, whether carryover was evaluated, or whether the analysis was based on participant-level paired data. The recalculations from arm summaries cannot answer these questions.

The primary endpoint wording also leaves uncertainty about whether minimum pH refers to the 1-minute value or the 7-minute value. The registry lists other measurement times and additional summaries, including area under the curve, but does not provide a clearly described analysis plan connecting those measures to the reported result. The registry record does not report whether the endpoint timing, contrasts, and multiplicity approach were prespecified. These gaps limit interpretation of the analysis beyond the results and labels posted.

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

The results describe measured dental biofilm pH under the dentifrice and sucrose-challenge conditions used in this study. The posted arm means range from 5.25 for fluoride to 5.7 for calcium glycerophosphate and fluoride. The sponsor reports a group-level ANOVA p-value of 0.001, while unadjusted recalculations give pairwise mean differences with intervals that include zero. These are separate analyses and should be interpreted as such.

The endpoint is a laboratory-style physiological measurement in dental biofilm, not a direct count of dental caries or an assessment of long-term clinical outcomes. The registry lists dental caries as a condition, but the primary measure is minimum pH after a sucrose challenge. The study’s posted data therefore do not establish effects on caries incidence, symptoms, or other outcomes not reported in the registry record. No medical advice or treatment recommendation follows from these results.

The study’s size and design also bound what can be inferred. It involved 12 young adults in a four-condition crossover study. The design allows participants to experience different dentifrices across periods, but the registry does not report the analysis details needed to assess how within-person dependence, period effects, or missing measurements were handled. The stated endpoint timing is not fully clear, and the sponsor’s confidence-interval limits are absent.

More generally, a statistically reported difference is not automatically a measure of clinical importance, and a non-significant pairwise p-value is not proof that conditions are equivalent. The recalculated intervals include a range of possible effects, while the sponsor’s ANOVA result addresses a group-level question under a model that is not described in sufficient detail. The evidence should therefore be understood as a report of pH measurements and analyses in this small study, not as a definitive answer to every question about the dentifrices or dental health.

16. Why This Trial Matters Statistically

This trial illustrates why the analysis question and the design must be considered together. A four-condition crossover study creates repeated measurements within participants, so a statistical method that recognizes that structure may differ from simple comparisons of arm-level means. The registry reports an ANOVA result, but does not describe its model. The unadjusted recalculations are useful for understanding the posted summaries, yet they cannot reconstruct a participant-level crossover analysis without paired data and the model details.

It also demonstrates the distinction between a global test and pairwise estimates. The sponsor’s ANOVA is reported across the four groups; the recalculations compare each active dentifrice with control. The two approaches address different questions and can produce results that are not directly comparable. Reporting an effect estimate with its interval is important because a p-value alone does not describe the range of effect sizes compatible with the data.

Endpoint definition is another statistical issue. The record calls the outcome minimum pH but gives a timeframe of 1 minute or 7 minutes, while the study summary describes repeated observations and area-under-the-curve summaries. A precise endpoint time and analysis rule are essential for interpreting which biological feature is being compared. The registry’s incomplete timing and model details make transparent interpretation harder, independent of the p-values.

Finally, planning calculations based on observed effects can illustrate how much information may be needed to estimate effects with chosen power targets. They do not reveal the original design rationale, which is not reported in the registry record, and they are not observed power. This distinction is especially important when the observed effects come from a small study and the planning assumptions simplify the crossover design.

17. 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 leaves several trial-specific limitations unresolved. The endpoint timeframe says “at 1 minute (minimum fermenting pH) or at 7 minutes,” so it is not clear which time point underlies the posted primary result. The summary describes multiple post-challenge measurements and additional measurement sets, but the record does not clearly connect those observations to the reported minimum pH value.

The study used a crossover design, but the registry does not state how the sponsor’s ANOVA modeled repeated measurements within participants, treatment period, sequence, or possible carryover. It describes a 1-week wash-out period but does not report evidence about whether that wash-out was sufficient. The posted arm summaries do not include paired participant-level data, which limits assessment of within-person variability and prevents reconstruction of a crossover analysis from the summary statistics alone.

The sponsor’s result is duplicated in the registry, gives a value of 5.5 under a parameter label of “Median Difference (Final Values),” and has no interval limits. The registry does not explain the statistic label, the duplicate entry, or the discrepancy between that report and the simple pairwise recalculations. The sponsor’s analysis takes precedence, but its model and full precision cannot be evaluated from the record.

Other details not reported in the registry record include a clear analysis population, missing-data handling, multiplicity procedures, adherence information, sequence-generation details, and a detailed statistical plan. The masking classification also differs between the design field, which states triple masking, and the brief summary, which says double blind; the record does not explain this difference. These limitations affect interpretation of the posted evidence but do not change the descriptive values reported.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Effect of Calcium Glycerophosphate (CaGP) - Fluoride Dentifrice on Dental Biofilm pH — NCT00875212.
• Registry record last updated 2009-10-15; status last verified 2009-10.
• Results first posted 2009-10-12.
• 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.

21. Record Summary

NCT00875212 is a completed Phase 1 randomized crossover study of dentifrices containing calcium glycerophosphate, fluoride, both, or neither. The study involved 12 young adults and measured dental biofilm pH after a sucrose challenge. The registry lists dental caries as the condition and gives minimum pH as the primary outcome.

The posted means are 5.4 for control, 5.48 for calcium glycerophosphate, 5.25 for fluoride, and 5.7 for the combination. The sponsor reports an ANOVA p-value of 0.001, while unadjusted recalculations give pairwise intervals that include zero. Because the registry flags a discrepancy, the sponsor’s analysis takes precedence. The registry does not report enough detail to resolve the model, endpoint timing, multiplicity, or missing-data questions.