Effect of Potassium Bicarbonate Supplementation on Bone and Muscle in Older Adults: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effect of Potassium Bicarbonate Supplementation on Bone and Muscle in Older Adults (NCT00357214), a randomized na study. The primary outcome is Biochemical Markers of Bone Turnover.

2. Clinical Question

The registry describes the rationale as investigating the role of dietary acid-base balance in bone and muscle health. Its brief summary proposes that excess acid may contribute to calcium loss and muscle breakdown, and identifies potassium bicarbonate as a base supplement. These statements explain the study motivation; they are not findings established by the posted outcome values.

The primary outcome is the three-month change in a 24-hour urine measure. The record provides group means and standard errors, along with unadjusted comparisons against potassium bicarbonate. It does not report a separate primary result for muscle wasting, and it lists no posted secondary outcomes. The clinical question answered by the reported numbers is therefore limited to this urinary marker over the stated timeframe.

3. Trial Design

The registry describes a randomized, triple-masked, parallel-group study. It lists four dietary-supplement arms: potassium bicarbonate, sodium bicarbonate, potassium chloride, and microcrystalline cellulose placebo. Each active supplement is described at 67.5 mmol/d. The record identifies Tufts University as sponsor and lists the study status as completed.

The registry gives a study start of 2006-09 and primary completion of 2008-04. It lists enrollment of 171. The outcome-measure counts differ by arm from the safety group counts, and the registry record does not explain the relationship between these reported counts. The group sizes attached to the primary outcome are 39 for potassium bicarbonate, 39 for sodium bicarbonate, 42 for potassium chloride, and 42 for microcrystalline cellulose.

The primary outcome is titled “Biochemical Markers of Bone Turnover.” It is the mean change in 24-hour urinary N-telopeptide/creatinine from baseline to three months, reported in nmol/mmol. The registry identifies the dispersion statistic as standard error. The comparisons posted are mean differences between potassium bicarbonate and each other arm.

Randomization and masking describe features of allocation and conduct; they do not, by themselves, establish how outcomes were analyzed or how missing observations were handled. The registry record does not report the original sample-size justification, detailed allocation procedures, masking roles, analysis populations, or a fuller statistical analysis plan. It lists no posted analyses, though outcome summaries and pairwise comparison statistics are available. No interim analyses are reported in the registry record.

The four-arm structure permits comparisons of the potassium bicarbonate arm with a bicarbonate comparator, a potassium comparator, and placebo. The posted contrasts are not a complete set of all possible pairwise comparisons among the four arms. Their interpretation should therefore stay tied to the specific contrasts and outcome that the registry reports.

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

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 Biochemical Markers of Bone Turnover, measured over 3 month change in 24-hr urine values.

The primary endpoint is change in 24-hour urinary N-telopeptide/creatinine from baseline to three months, expressed in nmol/mmol. N-telopeptide is a biochemical marker related to bone turnover; urinary creatinine is part of the reported ratio. A change endpoint describes how the measured value moved over the stated interval rather than only reporting a value at one time. The registry reports the mean change for each arm.

Biochemical markers can be used to assess biological processes relevant to bone, and may respond over a shorter timeframe than outcomes such as fractures. They are not interchangeable with those clinical outcomes. A change in this marker does not, on its own, establish a change in bone strength, fracture risk, muscle status, or how participants felt or functioned. Those outcomes are not the primary endpoint reported in the record.

The registry gives a standard error for each arm mean. A standard error describes the estimated precision of the mean, rather than the spread of individual participants' measurements. The record does not provide an individual-level distribution or a standard deviation for each arm. The endpoint is therefore best read as a comparison of average short-term marker changes, with uncertainty, and not as a complete account of bone or muscle health.

7. Results Posted in the Registry

• Sodium Bicarbonate: mean -3.36 versus -6.25 on control; difference 2.89 (95% CI -1.86 to 7.64), standardized effect d = 0.27; z-test p = 0.234.
• Potassium Chloride: mean -0.14 versus -6.25 on control; difference 6.11 (95% CI 1.31 to 10.91), standardized effect d = 0.56; z-test p = 0.013.
• Microcrystalline Cellulose: mean 0.50 versus -6.25 on control; difference 6.75 (95% CI 2.09 to 11.41), standardized effect d = 0.63; z-test p = 0.005.

8. Statistical Interpretation

9. Statistical Methods Explained

For a continuous outcome such as the change in urinary N-telopeptide/creatinine, the mean summarizes the average observed change within an arm. The means in the registry are -6.25 for potassium bicarbonate, -3.36 for sodium bicarbonate, -0.14 for potassium chloride, and 0.5 for microcrystalline cellulose. Negative values represent decreases from baseline on the reported change scale, while a positive value represents an increase. This describes the measurement direction; it does not imply that any direction is necessarily better for participants.

The standard error (SE) describes the precision of an estimated mean. It is not the same as the standard deviation, which describes variation among individual observations. The reported SE values are 1.75, 1.68, 1.71, and 1.61 for the four arms, respectively. A smaller SE indicates a more precisely estimated mean, considered in relation to the scale and measurement. It does not mean that individual responses are less variable.

What does a mean difference represent?

A mean difference compares the average outcome in two arms. For these posted contrasts, the listed estimate is the comparison-arm mean minus the potassium bicarbonate-arm mean. The sodium bicarbonate contrast is 2.89; the potassium chloride contrast is 6.11; and the microcrystalline cellulose contrast is 6.75. A value of zero would correspond to equal arm means. Positive estimates indicate higher mean change in the named comparison arm than in potassium bicarbonate on this endpoint. The estimate is a contrast in group averages, not a description of how every participant responded.

The comparisons are unadjusted recalculations from the posted summary statistics. The registry does not list formal analyses, despite posting these comparison statistics. It also does not report whether a different, prespecified model was used by the study team. The registry record does not report covariate adjustment, handling of repeated measurements, or a formal analysis population. Accordingly, the contrasts should not be described as adjusted trial-model estimates.

How should p-values be read?

The unadjusted p-value assesses how unusual a result at least as far from the null as the observed result would be if the null hypothesis of no mean difference were true, under the test's assumptions. The posted values are 0.234 for sodium bicarbonate, 0.013 for potassium chloride, and 0.005 for microcrystalline cellulose. A p-value does not give the probability that the null is true, the probability that a finding will replicate, or the clinical value of an effect. It is also sensitive to the amount of information and the analysis assumptions.

Confidence intervals complement p-values by displaying an estimated range for the contrast. A 95% confidence interval is produced by a procedure that, under its assumptions, would cover the target value in 95% of repeated applications. It is not a guarantee that a particular interval contains the true value. The confidence intervals for potassium chloride and microcrystalline cellulose are above zero, while the sodium bicarbonate interval includes zero. These statements describe the posted unadjusted intervals, not multiplicity-adjusted inference.

Why does multiplicity matter?

When several hypotheses are tested, the chance of at least one false-positive result across the set can be greater than for an individual test. The record posts three comparisons against potassium bicarbonate, but it does not report a multiplicity procedure or a family-wise testing strategy. The unadjusted p-values should therefore be understood as individual-comparison statistics, not as results corrected for examining multiple contrasts. The record also does not say whether these comparisons were the only hypotheses considered during the study.

What cannot be reconstructed from group summaries?

Summary means and standard errors do not disclose the full distribution of measurements, possible outliers, baseline balance, participant-level changes, or the reasons for missing outcome measurements. Nor do they allow independent evaluation of model diagnostics or the handling of protocol deviations. The registry record does not report these details. The posted group summaries are informative about average changes and their estimated precision, but they cannot replace a complete analysis report or participant-level data.

General statistical context: an assumption of normally distributed estimates is often used for mean comparisons, especially when applying standard confidence-interval procedures. Whether the trial's original analysis used that approach, a different model, or additional adjustments is not reported in the registry record. The simple comparisons should not be mistaken for confirmation of the trial's full statistical plan.

10. Confidence Intervals and Precision

The confidence intervals describe uncertainty around each mean difference, under the assumptions of the comparison procedure. For sodium bicarbonate versus potassium bicarbonate, the estimated difference is 2.89 and the 95% confidence interval runs from -1.865 to 7.645. Because that range includes zero, the data are compatible with no mean difference as well as with contrasts on either side of zero. The interval is also compatible with positive differences of varying size.

For potassium chloride versus potassium bicarbonate, the estimated difference is 6.11 and the 95% confidence interval is 1.314 to 10.91. For microcrystalline cellulose versus potassium bicarbonate, the estimate is 6.75 and its 95% confidence interval is 2.089 to 11.41. Both intervals are entirely above zero. Their widths indicate that the point estimates should not be read as exact values; a range of contrasts remains compatible with the posted results and the assumptions used.

Precision concerns the range of plausible values, not whether a result matters clinically. The registry record does not report a threshold for a meaningful difference in urinary N-telopeptide/creatinine, and it does not connect this marker contrast to a direct clinical outcome. Therefore, the intervals cannot establish clinical importance on their own.

The intervals and p-values are unadjusted. They do not reflect a reported correction for multiple comparisons, and the registry record does not supply details of the original analysis model. In addition, group means and standard errors do not reveal individual-level variability or the effects of missing observations. These limitations qualify the apparent precision of the contrasts as estimates for the reported endpoint.

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 sample-size calculations are planning illustrations based on the observed effect, not the trial's own design calculations. For these continuous-outcome contrasts, the listed figures are participants per arm, not total event counts. The assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect defined as the stated fraction of the observed effect.

For each contrast, the 80% power column gives the number of participants per arm estimated to yield an 80% chance of detecting the selected effect under those assumptions, if that effect is present. The 90% power column gives the corresponding planning number for a 90% chance. These are planning probabilities under a model, not guarantees for a future study. The calculations use the observed contrast as their starting point, so they should not be treated as independent confirmation of the trial findings.

The rows for 100%, 75%, and 50% of the observed effect illustrate how planning changes when the target effect is smaller than the observed contrast. Smaller effects need more participants for a given power target, which is why the participant-per-arm figures increase as the target fraction decreases. Each comparison has its own values because the observed contrasts and estimated variability differ.

The curve should be read as a visual companion to those planning values under the same stated assumptions, rather than as an additional trial result. The registry record does not report the original sample-size justification, the assumptions used when designing the trial, or whether any planned enrollment target was based on these calculations. 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 = 2.890; pooled standard deviation σ = 10.712.
2. n = 2 × (1.96 + 0.842)² × 10.712² / 2.890² = 215.7, rounded up to 216 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 <- 2.89                      # observed difference in means
sigma <- 10.7124016914976                     # 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(357214)
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 = 216 (80% power), 289 (90% power)
 75% of observed effect: n per arm = 384 (80% power), 514 (90% power)
 50% of observed effect: n per arm = 863 (80% power), 1,155 (90% power)
power.t.test: 216.6 
Simulated power at n = 216 per arm: 0.800 (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. Multiplicity, Missing Data and Analysis Populations

The registry posts three comparisons of potassium bicarbonate with the other arms, each with an unadjusted p-value and confidence interval. It does not report a multiplicity adjustment, a hierarchy of hypotheses, or a family-wise testing plan. Since several comparisons are available, the unadjusted values do not account for the possibility of false-positive findings across the full set. This is an important qualification when interpreting the contrast-level statistics.

The record lists one primary outcome measure and no posted secondary outcomes. It does not state whether additional outcomes or analyses were measured but not reported, or whether the posted contrasts were all prespecified. The number of analyses listed is zero, while comparison statistics are nevertheless posted. A detailed analysis plan is not reported in the registry record, so the relationship between the reported summaries and the study's original analysis plan is unclear.

Missing-data handling is also not reported in the registry record. The endpoint entries list group sizes of 39, 39, 42, and 42; safety risk groups are listed separately as 43, 43, 42, and 43. The record does not explain why the counts differ or whether the outcome summaries include every randomized participant. It does not report the extent or reasons for missing endpoint values, imputation methods, or sensitivity analyses. No assumption about missingness can be verified from the posted summaries.

The registry record does not identify an intention-to-treat, per-protocol, or other analysis population for the primary endpoint. Randomization helps define the design but does not establish which participants were included in each analysis. Nor does the record describe exclusions after randomization, adherence, or protocol deviations. Those omissions limit assessment of whether the reported group summaries preserve the randomized comparison.

These gaps do not change the arithmetic of the posted summaries, but they limit how completely readers can assess bias and inferential uncertainty. The unadjusted comparisons should be interpreted as comparisons based on the posted aggregate statistics, rather than as a full account of the original analysis process.

14. 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 0% reporting threshold.

Most frequent non-serious adverse event terms

The registry lists serious adverse effects as affecting zero participants in each safety group. The corresponding numbers at risk are 43 for potassium bicarbonate, 43 for sodium bicarbonate, 42 for potassium chloride, and 43 for microcrystalline cellulose. These are the reported group counts for the serious adverse-effect measure. The record lists no serious adverse-effect terms.

These figures are a descriptive safety summary, not a statistical comparison between arms. A zero reported count does not show that an event is impossible, and the record does not provide confidence intervals or a test comparing groups. It also does not describe event ascertainment or provide further context for interpreting the observation period.

The death counts are null for all arms. Death information is therefore not reported in the registry record; a null entry should not be read as a confirmed zero. Other adverse effects are listed separately: one for potassium bicarbonate, three for sodium bicarbonate, seven for potassium chloride, and one for microcrystalline cellulose, with the same group-specific risk counts. These figures should not be confused with the serious adverse-effect counts.

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

The posted endpoint results show average changes in urinary N-telopeptide/creatinine over three months and unadjusted mean differences between potassium bicarbonate and the other trial arms. The potassium bicarbonate mean change was -6.25 nmol/mmol; the other arm means were -3.36, -0.14, and 0.5. The posted contrasts and their confidence intervals provide estimates of how those arm averages differed on this particular marker.

The comparisons with potassium chloride and microcrystalline cellulose have confidence intervals above zero, while the sodium bicarbonate comparison has an interval that includes zero. These are observations about the posted estimates and their unadjusted uncertainty. They are not, by themselves, evidence that a difference is clinically important. The registry gives no clinical threshold for this marker, and the endpoint is not a direct measure of fracture occurrence, bone strength, muscle wasting, muscle function, or quality of life.

The study rationale refers to possible effects of dietary acid-base balance on bone and muscle. The available results do not establish all parts of that rationale. In particular, the registry record does not post secondary outcome results, and it does not report longer-term outcomes. A short-term biomarker comparison cannot answer questions that require other measures or a different timeframe.

The results also need to be read with uncertainty about multiplicity, missing data, analysis populations, and the original statistical plan. The record does not report adjustment for multiple comparisons or explain the difference between endpoint and safety group counts. These limitations do not erase the reported values, but they constrain conclusions that can be drawn from them.

Accordingly, the most defensible reading is narrow: the registry reports group-average changes in a biochemical marker and unadjusted contrasts against potassium bicarbonate, with the stated confidence intervals. The results alone do not establish a treatment recommendation or prove an effect on broader health outcomes.

16. Why This Trial Matters Statistically

This trial illustrates why the endpoint and the clinical question must be kept distinct. The study motivation concerns bone and muscle health, while the posted primary outcome is a urinary biochemical marker measured as change from baseline to three months. Statistical differences in a marker can be informative about the measured process, but they do not automatically translate into differences in fractures, muscle function, or other outcomes that were not reported.

The four-arm design also shows the importance of identifying the exact comparison. The registry reports contrasts of potassium bicarbonate with sodium bicarbonate, potassium chloride, and microcrystalline cellulose. The direction and uncertainty of each contrast differ, and the results should not be collapsed into a single general statement about all supplements. The unadjusted p-values and confidence intervals answer questions about individual contrasts, while the record does not describe how multiple testing was handled.

Precision matters alongside the point estimates. Confidence intervals communicate how uncertain the estimated differences are, and the sodium bicarbonate interval includes zero while the other posted intervals do not. Even intervals that exclude zero do not establish clinical relevance without an interpretation threshold or a linked clinical outcome. The standard errors and confidence intervals thus help separate evidence about a group average from certainty about its magnitude and importance.

Finally, the registry record offers a useful but incomplete statistical account. It posts outcome summaries and comparisons but does not report a full analysis plan, missing-data approach, or analysis population. Recognizing those omissions is part of interpreting the evidence responsibly. The case demonstrates how careful statistical reading can preserve what the data show while avoiding conclusions beyond the measured endpoint.

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 primary outcome is a short-term urinary marker, and the posted results do not include direct measures of fractures, bone strength, muscle wasting, muscle performance, or participant-reported outcomes. The registry lists no secondary outcomes. Thus, the outcome summaries cannot establish whether the observed marker changes correspond to other aspects of bone or muscle health.

The group counts reported for the endpoint do not match all of the safety risk-group counts. The registry record does not explain this difference, identify the endpoint analysis population, or describe missing observations. Reasons for missing data, imputation, and sensitivity analyses are not reported in the registry record.

The comparisons are unadjusted, and no multiplicity procedure is reported despite multiple posted contrasts. The record also does not provide the original sample-size justification, a detailed statistical analysis plan, interim-analysis information, covariate-adjustment details, or model diagnostics. It is therefore not possible to establish from the registry summaries whether the posted calculations reflect all planned analyses or how analysis decisions were made.

The registry supplies group means and standard errors rather than individual-level data or full distributions. This restricts evaluation of outliers, variability, baseline patterns, and assumptions behind the comparisons. The confidence intervals quantify uncertainty under the comparison procedure but do not account for unreported design or analysis choices.

These limitations are specific to the available record and endpoint; they do not imply that unreported procedures were or were not undertaken. Where such details are absent, they are not reported in the registry record.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Effect of Potassium Bicarbonate Supplementation on Bone and Muscle in Older Adults — NCT00357214.
• Registry record last updated 2020-02-17; status last verified 2020-02.
• Results first posted 2020-02-05.
• 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

NCT00357214 is a completed randomized, triple-masked, parallel-group trial sponsored by Tufts University. The registry lists enrollment of 171 and a study start of 2006-09, with primary completion in 2008-04. The four arms received potassium bicarbonate, sodium bicarbonate, potassium chloride, or microcrystalline cellulose placebo.

The primary outcome is mean change in 24-hour urinary N-telopeptide/creatinine from baseline to three months, in nmol/mmol. The registry reports arm means, standard errors, and three unadjusted comparisons against potassium bicarbonate. It lists one outcome measure and no posted secondary outcomes. The record does not report the original sample-size justification, detailed analysis plan, missing-data handling, or analysis population.

The results describe short-term changes in a biochemical marker. They should not be treated as direct evidence about fractures, muscle outcomes, or clinical importance, and the posted contrasts do not include a reported multiplicity adjustment.