The Effect of Boosting Dietary Lactobacillus and Phytochemical Rich Foods on Biomarkers of Longevity—A Phase II Randomised Placebo Controlled Trial

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The topic is relevant to the scope of the Journal of Aging and Longevity, particularly given the growing interest in modifiable nutritional strategies targeting functional and inflammatory biomarkers of aging. The randomized design, relatively large sample size, and focus on clinically significant outcomes, such as grip strength, are clear strengths. However, several methodological and interpretative issues need to be addressed before the manuscript can be considered for publication. The considerations outlined below aim to reinforce the scientific rigor and clarity of the work.

1. Study design and interpretation of effects

• All participants received the PRS, and randomization was limited to probiotic versus placebo. Therefore, the study design does not allow for causal inferences regarding the independent effect of the phytochemical supplement itself.
• While improvements in grip strength were observed in both arms, these changes cannot be definitively attributed to the PRS in the absence of a true control group. Factors such as test-retest effects, behavioral changes due to trial participation, or regression to the mean could have contributed.The manuscript would benefit from clearer phrasing in both the Results and Conclusions sections to indicate that the randomized comparison primarily evaluates the additional effect of the probiotic on a PRS background, rather than the isolated effect of phytochemical supplementation.

 

1. Baseline Imbalance in grip strength

• Baseline grip strength differed notably between groups (PRS+P: 31.1 kg vs PRS+PB: 36.4 kg). This difference is clinically relevant and warrants more explicit consideration.
• Although changes over time are reported, it is not entirely clear whether the statistical approach adequately adjusted for baseline differences. An ANCOVA model adjusting for baseline grip strength would be more appropriate than relying solely on change scores or mixed ANOVA without explicit clarification. Please clarify:
• Whether baseline grip strength was included as a covariate.
• Whether between-group comparisons of change were adjusted accordingly.
• The magnitude of effect size (e.g., Cohen’s d) for the between-group difference.

 

1. Testosterone Analysis

• There are inconsistencies in the reporting of testosterone results:
• Units alternate between ng/dL and ng/L.
• The reported standard deviation in one group appears implausible (“± 476 ng/L”), likely a typographical error.
• Baseline testosterone values are not presented.
• It is unclear whether changes from baseline were analyzed or whether only 4-month values were compared.
• Given that testosterone is a central outcome in the manuscript’s interpretation, these issues should be corrected and clarified. In addition, a brief discussion of the clinical relevance of the observed difference (beyond statistical significance) would strengthen the manuscript.

 

1. Inflammatory marker selection

• NLR was used as the sole marker of systemic inflammation. While NLR is practical and inexpensive, it is a relatively nonspecific marker. The absence of additional inflammatory biomarkers (for example, CRP, IL-6, TNF-α) should be more clearly acknowledged as a limitation.
• The discussion section could be strengthened by elaborating on the limitations of NLR and the need for more comprehensive inflammatory profiling in future studies.

 

The study addresses an important and timely topic within ageing research. With clarification of statistical methods, correction of reporting inconsistencies, more cautious interpretation, and improved discussion of limitations, the manuscript has the potential to make a valuable contribution.

Author Response

Reviewer 1:

Thank you for your very thorough and detailed review of this study. Your suggestions are welcome and they will make the paper more robust and academically appropriate. I have worked with the other authors and statistician to make all the appropriate changes in document.

Comment.1 Study design and interpretation of effects

• All participants received the PRS, and randomization was limited to probiotic versus placebo. Therefore, the study design does not allow for causal inferences regarding the independent effect of the phytochemical supplement itself.
• While improvements in grip strength were observed in both arms, these changes cannot be definitively attributed to the PRS in the absence of a true control group. Factors such as test-retest effects, behavioural changes due to trial participation, or regression to the mean could have contributed. The manuscript would benefit from clearer phrasing in both the Results and Conclusions sections to indicate that the randomized comparison primarily evaluates the additional effect of the probiotic on a PRS background, rather than the isolated effect of phytochemical supplementation.

Answer: Thank you - This was the phase II element of the study which as you correctly point out, as in all published phase II trials, in view of the absence of a placebo control could have been a natural variability or even a placebo effect. Although, phase II data can be interesting, are reported widely in the literature and can direct further research, I completely agree they are less convincing -  We have now re-phrased and highlighted, that this observation this requires further investigation to be substantiated. We have also altered the conclusions accordingly. An explanation for this aspect of the design is included in the discussion. 

Comment 2 Baseline Imbalance in grip strength

• Baseline grip strength differed notably between groups (PRS+P: 31.1 kg vs PRS+PB: 36.4 kg). This difference is clinically relevant and warrants more explicit consideration.

Answer: Agreed - we have now highlighted, in more detail, the differences in the baseline characteristics and added these to table 1 in order to emphasized them to the reader. We have added that men in the PRS+PB at baseline were, by chance, slightly younger and stronger. The implications have been added to the discussion and a line outlining this difference has been added to the relevant section of the results.

Comment 3

• Although changes over time are reported, it is not entirely clear whether the statistical approach adequately adjusted for baseline differences. An ANCOVA model adjusting for baseline grip strength would be more appropriate than relying solely on change scores or mixed ANOVA without explicit clarification. Please clarify:
  • Whether baseline grip strength was included as a covariate.
  • Whether between-group comparisons of change were adjusted accordingly.
  • The magnitude of effect size (e.g., Cohen’s d) for the between-group difference.

Answer 3. As regards the statistical analysis a linear mixed model has been performed to evaluate the difference between the two groups. The fixed factors (arm and time) and random (participant) effects for each LMM, were used. The grip strength at baseline was used as a covariate. Using the smallest Hurvich and Tsai criterion (AICC) an appropriate model was chosen for each variable. This type of analysis was preferred as it allows for missing data, can accurately model different covariate structures for repeated measures data and can model between-subject variability. Step down Hommel adjusted post-hoc pairwise comparisons were calculated if a significant main effect and/or interaction effect was present. Where a significant main and / or interaction effect was obtained a Sidak post-hoc tests were used to locate significant differences.

Information regarding the regression analysis was omitted from the paper - in view of the word count- but this has now been added. Linear regression was also performed on % PSA change from baseline to 4 months with mean grip strength difference over the same time period. Prior to the regression analyses, continuous predictors were assessed for multicollinearity using variance inflation factors (VIF), and categorical variables were dummy-coded where necessary. The Hos-mer-Lemeshow goodness-of-fit test was used to evaluate the model's fit, and Nagelkerke’s R² was reported to estimate the proportion of variance explained by the model. Odds ratios (OR) with 95% confidence intervals (CI) were calculated for each pre-dictor to quantify the strength and direction of associations with the outcome variable.

For the main end point, the "changes" from baseline at 4 months  were compared between the two randomised groups - not the absolute levels at 4 months so we believe difference in baseline ages are unlikely to affect this comparison which was statistically significant. We also have conducted a literature search to try and establish whether the differences in the baseline levels could have influences the "change difference" between the two groups - i.e. are older, stronger men more likely to have an improvements in GS following an intervention compared to younger weaker men - we found no literature to suggest this

 

Comment 4: Testosterone Analysis

• There are inconsistencies in the reporting of testosterone results:
  • Units alternate between ng/dL and ng/L - thank you for highlighting these errors, after going back to the source data on the hospital results system we have corrected to nmol/l
  • The reported standard deviation in one group appears implausible (“± 476 ng/L”), likely a typographical error. - corrected
  • Baseline testosterone values are not presented. It is unclear whether changes from baseline were analyzed or whether only 4-month values were compared.

Answer 4. yes - We were unable to obtain testosterone levels at baseline which I agree would been useful but this would have involved sending men back for an additional blood test and advise from our patient advocate panels was this would have been a barrier to recruitment

• Given that testosterone is a central outcome in the manuscript’s interpretation, these issues should be corrected and clarified. In addition, a brief discussion of the clinical relevance of the observed difference (beyond statistical significance) would strengthen the manuscript.

 Answer - This has been added - thank you

1. Inflammatory marker selection

• NLR was used as the sole marker of systemic inflammation. While NLR is practical and inexpensive, it is a relatively nonspecific marker. The absence of additional inflammatory biomarkers (for example, CRP, IL-6, TNF-α) should be more clearly acknowledged as a limitation.
• The discussion section could be strengthened by elaborating on the limitations of NLR and the need for more comprehensive inflammatory profiling in future studies.

Answer: Agreed - although clinically we find this a useful tool (mainly as you say it's cheap and easily available) for a clinical study more complex markers would be better - I have added this to the conclusion.

 

Reviewer 2 Report

Comments and Suggestions for Authors

The authors conducted a RCT targeting an elderly population, but there was a significant age difference between the two groups, with average ages of 76 and 73 years old. Although both groups were in their 70s, an older age by several years can lead to notable differences in muscle condition and response to intervention, thereby weakening the persuasiveness of the results. For example, this difference may have contributed to the disparity in baseline grip strength—31.1 kg in the older group and 36.4 kg in the younger group—suggesting that the baseline characteristics of the two groups may not be comparable. The authors attempted to assess the intervention effect by examining the improvement in grip strength from baseline to four months post-intervention. However, due to differences in age and baseline muscle status, the extent of improvement over this period may naturally vary, making it difficult to attribute any significant difference solely to the intervention. Furthermore, although the authors aimed to investigate the combined effect of Dietary Lactobacillus and Phytochemicals-Rich Foods, since the control group also received Phytochemicals-Rich Foods, the main variable essentially became Dietary Lactobacillus. This makes it challenging to determine whether the observed effects truly stem from the combined intervention. Thus, the study design has limitations. Minimizing these limitations—for instance, by excluding extreme age outliers to enhance comparability between groups—could lead to more robust conclusions. Additionally, the formatting of p-values in the text should be standardized, as they are inconsistently presented with variations in capitalization (uppercase/lowercase) and italicization.

Author Response

Thank you for your very thorough and detailed review of this study. Your suggestions are welcome and they will make the paper more robust and academically appropriate. I have worked with the other authors and statistician to make all the appropriate changes in document.

Comment 1 The authors conducted a RCT targeting an elderly population, but there was a significant age difference between the two groups, with average ages of 76 and 73 years old. Although both groups were in their 70s, an older age by several years can lead to notable differences in muscle condition and response to intervention, thereby weakening the persuasiveness of the results. For example, this difference may have contributed to the disparity in baseline grip strength—31.1 kg in the older group and 36.4 kg in the younger group—suggesting that the baseline characteristics of the two groups may not be comparable.

Answer Agreed - we have now highlighted, in more detail, the differences in the baseline characteristics and added these to table 1 in order to emphasized them to the reader. We have added that men in the prv+PB at baseline were, by chance, slightly younger and stronger.

In the main analysis from GS the "changes" from baseline at 4 months  were compared between the two randomised groups - not the absolute levels at 4 months so we believe difference in baseline ages are unlikely to affect this comparison. Also the statistician had adjusted for variations in baseline characteristics (now highlighted in the methodology). That said, we have conducted a literature search to try and establish whether the differences in the baseline age and brip could have influenced the level of changes between the two randomised groups at 4 months - i.e. are younger men, stronger men more likely to have an greater improvements in GS following an intervention compared to younger weaker men. There is no literature to suggest this which would support our clinical impression.

Comment: The authors attempted to assess the intervention effect by examining the improvement in grip strength from baseline to four months post-intervention. However, due to differences in age and baseline muscle status, the extent of improvement over this period may naturally vary, making it difficult to attribute any significant difference solely to the intervention.

Answer: We agree that changes in outcomes between randomised arm could have happened by chance but we feel the study was adequately powered and with a p value of <0.001 this is very likely to be a genuine effect of the intervention

 

Comment: Furthermore, although the authors aimed to investigate the combined effect of Dietary Lactobacillus and Phytochemicals-Rich Foods, since the control group also received Phytochemicals-Rich Foods, the main variable essentially became Dietary Lactobacillus.

Answer: Agreed - the difference in the two arms was placebo or PB, both on a background of the PRS - The patient advocacy groups, advised that a placebo only group would have been unacceptable for them, so we would not have recruited - hence the compromise in terms of a PRS alone versus placebo alone arm. However, as men we not taking another other supplements and there was a clear difference between the two blinded randomised arms we believe this is statistically robust and there are genuine differences between men on PRS alone versus the PRS+ PB.

As regards the  phase II element of the study as highlighted in the methodology. As you point out,  as in all published phase II trials, in view of the absence of a placebo control the "before and after" comparison could have been a natural variability or even a placebo effect. Although, phase II data can be interesting and can direct further research, I completely agree they are less statistically robust -  I have now re-phrased and highlighted, that this observation this requires further investigation to be substantiated. We have also altered the conclusions accordingly. An explanation for this aspect of the design is included in the discussion. 

Comment: This makes it challenging to determine whether the observed effects truly stem from the combined intervention. Thus, the study design has limitations. Minimizing these limitations—for instance, by excluding extreme age outliers to enhance comparability between groups—could lead to more robust conclusions.

Answer: Thank you and we agree but as we did not pre-determine this in the statistical plan we felt it would be more appropriate to use the complete data

Comment: Additionally, the formatting of p-values in the text should be standardized, as they are inconsistently presented with variations in capitalization (uppercase/lowercase) and italicization.

Answer: Thank you - these have been corrected

Reviewer 3 Report

Comments and Suggestions for Authors

In this review, Thomas and coll. presents results from a phase II double-blind randomized placebo-controlled trial investigating the combined effects of a phytochemical-rich supplement (PRS) and a multi-strain Lactobacillus probiotic on grip strength, testosterone levels, and neutrophil-to-lymphocyte ratio (NLR) in older men undergoing active surveillance for early prostate cancer.

The topic is timely and relevant given increasing interest in nutritional modulation of aging-related biomarkers and gut-muscle-endocrine interactions. The study benefits from a randomized, controlled design (for the probiotic component), ethical approval, trial registration, and an adequate sample size based on an a priori power calculation. However, significant methodological and statistical issues currently limit the strength of the conclusions.

Major Comments

Page 1, lines 32; page 2, line 78; page 7, lines 272-273. The manuscript repeatedly refers to “surrogate markers of longevity.” However, the study duration was only 4 months, surrogate biomarkers alone do not establish longevity effects, and no hard clinical endpoints were assessed. Therefore, the language should be moderated to avoid overstating implications.

Page 4, lines 122-123. There is a statistically significant age difference between groups (76 vs 73 years, p = 0.02). As age is strongly associated with grip strength, testosterone levels, and inflammatory markers, this imbalance may confound the primary outcome. In addition, baseline grip strength differs substantially between groups (page 5, Table 2: 31.1 kg vs 36.4 kg). It is unclear whether this difference was statistically significant, and no adjustment appears to have been performed. Therefore, the authors should adjust for age and baseline grip strength and explicitly report baseline between-group statistical comparisons. Without appropriate adjustment, it is difficult to conclude that the additional improvement observed in the probiotic arm is attributable solely to the intervention.

Page 4, lines 126-128. All participants received the phytochemical-rich supplement (PRS), and only the probiotic component was randomized. Therefore, the trial evaluates: PRS + Probiotic versus PRS + Placebo. It does not allow assessment of PRS versus placebo alone. Statements suggesting that “boosting dietary phytochemicals was associated with improved grip strength” (page 1, line 29) are not supported by a parallel placebo-only arm. Improvements over time may reflect natural variability, regression to the mean, behavioral changes, or study participation effects. The interpretation should be reframed to reflect the actual comparison.

Page 6, lines 224-225 and following. Although statistical significance is reported, clinical significance is not sufficiently discussed. Is a 2.5–4.4 kg increase in grip strength over 4 months clinically meaningful in this age group? Is a 1.73-unit increase in testosterone biologically relevant? What is the clinical implication of a 0.79 change in NLR? In addition, important methodological details regarding testosterone measurement are missing (time of blood sampling, fasting status, assay method used, and whether total or free testosterone was assessed). Given the known diurnal variation of testosterone, these details are critical.

Page 7, lines 237-241. The reported increase in testosterone in the probiotic arm is interpreted as a positive outcome in the context of aging biomarkers. However, given that the study population consists of men with prostate cancer, and considering the androgen sensitivity of prostate tumors, the clinical implications of increased testosterone warrant careful discussion. The manuscript should clarify whether testosterone levels remained within the physiological range, provide oncologic safety data (e.g., PSA kinetics or progression outcomes), and discuss the potential risks of increasing androgen levels in this population.

Minor Comments

Numerous typographical and grammatical errors require professional editing.

Page 1, line 19. Abstract. The capital letter in “Hundred” is unnecessary. Line 23 please replace “art” with “at”.

Page 3, line 119, page 4, line 120. Please revise grammar.

Page 4, line 136 and following. The Latin names of plants should be written with a capital letter for the genus and a lowercase letter for the species.

Page 5, line 173. Please correct “Kolmogorov”.

Page 6, table 3. Testosterone units are reported inconsistently (ng/dL vs ng/L) and the standard deviation for testosterone in the PRS+PB group (“± 476”) appears erroneous. Please amend.

Page 7, line 234. Please verify whether “effect” and “trail” are correct (likely “affect” and “trial”).

 

Author Response

Thank you for your very thorough and detailed review of this study. Your suggestions are welcome and they will make the paper more robust and academically appropriate. I have worked with the other authors and statistician to make all the appropriate changes in document.

Comment: Page 1, lines 32; page 2, line 78; page 7, lines 272-273. The manuscript repeatedly refers to “surrogate markers of longevity.” However, the study duration was only 4 months, surrogate biomarkers alone do not establish longevity effects, and no hard clinical endpoints were assessed. Therefore, the language should be moderated to avoid overstating implications.

Answer: Agreed - we have changed all these sections as suggested to avoid overstating the findings. We have also re-enforced this was a 4 months intervention which gives guidance for confirmatory future studies 

Comment: Page 4, lines 122-123. There is a statistically significant age difference between groups (76 vs 73 years, p = 0.02). As age is strongly associated with grip strength, testosterone levels, and inflammatory markers, this imbalance may confound the primary outcome.

Answer: As regards testosterone - As testosterone was only measured at 4 months, we completely agree the difference between the two groups, although not certain, could also  have been simply because of their age difference. We have therefore removed testosterone as a  conclusion in the abstract (to de-emphasise it's relevance and explained this clearly in the discussion.

As regards GS and inflammation (NLR) - the "changes" from baseline at 4 months  were compared between the two randomised groups - not the absolute levels at 4 months so we believe difference in baseline ages are unlikely to affect this comparison. That said, we have conducted a literature search to try and establish whether the differences in the baseline levels could have influenced the observed changes between the two randomised groups at 4 months - i.e. are younger men, stronger men more likely to have an greater improvements in GS following an intervention compared to younger weaker men There is no literature to suggest this which would support our clinical impression

Comment:  Baseline grip strength differs substantially between groups (page 5, Table 2: 31.1 kg vs 36.4 kg).

Answer: Thank you for highlighting this - We have highlighted this point by adding Grip Strength as a baseline demographic in table 1 with statistical parameters and made a comment in the text.

Comment: It is unclear whether this difference was statistically significant, and no adjustment appears to have been performed.  Therefore, the authors should adjust for age and baseline grip strength and explicitly report baseline between-group statistical comparisons. Without appropriate adjustment, it is difficult to conclude that the additional improvement observed in the probiotic arm is attributable solely to the intervention.

Answer: Thank you - The statistician did actually correct for baseline differences and the methodology for this has now been included in statistics section of the paper.

Comment: Page 4, lines 126-128. All participants received the phytochemical-rich supplement (PRS), and only the probiotic component was randomized. Therefore, the trial evaluates: PRS + Probiotic versus PRS + Placebo. It does not allow assessment of PRS versus placebo alone. Statements suggesting that “boosting dietary phytochemicals was associated with improved grip strength” (page 1, line 29) are not supported by a parallel placebo-only arm. Improvements over time may reflect natural variability, regression to the mean, behavioural changes, or study participation effects. The interpretation should be reframed to reflect the actual comparison.

Answer: This was the phase II element of the study as highlighted in the methodology. As you point out,  as in all published phase II trials, in view of the absence of a placebo control the "before and after" comparison could have been a natural variability or even a placebo effect. Although, phase II data can be interesting, can direct further research and are reported widely, we completely agree they are less statistically robust -  We have, therefore now re-phrased and highlighted, that this observation this requires further investigation to be substantiated. We have also altered the conclusions accordingly. An explanation for this aspect of the design is included in the discussion. 

Comment: Page 6, lines 224-225 and following. Although statistical significance is reported, clinical significance is not sufficiently discussed. Is a 2.5–4.4 kg increase in grip strength over 4 months clinically meaningful in this age group?

Answer: Agreed, but hard to qualify this and we don't want to overstate the result, just report them. Anecdotally men in the study often reported they felt better but formal correlation with QOL scores or activities of daily living were not done - future trials should certainly include these - we have already highlighted in the introduction the general importance of GS and we have added, following your suggestion, in the discussion a statement related to its clinical relevance and guidance for future studies

Comment: Is a 1.73-unit increase in testosterone biologically relevant? What is the clinical implication of a 0.79 change in NLR? In addition, important methodological details regarding testosterone measurement are missing (time of blood sampling, fasting status, assay method used, and whether total or free testosterone was assessed). Given the known diurnal variation of testosterone, these details are critical.

Answer: Agreed this was a small difference and levels remained in the normal range in both arms. The main caveat was that testosterone was measured at the end of the trial only and it is possible that these differences could be a results of the slight differences in baseline age demographics between the two randomised groups. We have added that further trials should ensure testosterone is measured at baseline and end of the intervention and included more detailed measurement such as free and bound levels as well as correlation with QOL scores. 

Comment: Page 7, lines 237-241. The reported increase in testosterone in the probiotic arm is interpreted as a positive outcome in the context of aging biomarkers. However, given that the study population consists of men with prostate cancer, and considering the androgen sensitivity of prostate tumors, the clinical implications of increased testosterone warrant careful discussion. The manuscript should clarify whether testosterone levels remained within the physiological range, provide oncologic safety data (e.g., PSA kinetics or progression outcomes), and discuss the potential risks of increasing androgen levels in this population.

Answer: Agreed - the effect on PSA/MRI was reported in a separate analysis  and publication - We wanted to avoid overstating in the conclusions of this study but, at your suggestion, I have added a sentence reassuring readers that this small rise in testosterone did not adversely affect men's risk of their prostate cancer progression

 

Minor Comments

Numerous typographical and grammatical errors require professional editing.

Page 1, line 19. Abstract. The capital letter in “Hundred” is unnecessary. Line 23 please replace “art” with “at”.

Page 3, line 119, page 4, line 120. Please revise grammar.

Page 4, line 136 and following. The Latin names of plants should be written with a capital letter for the genus and a lowercase letter for the species.

Page 5, line 173. Please correct “Kolmogorov”.

Page 6, table 3. Testosterone units are reported inconsistently (ng/dL vs ng/L) and the standard deviation for testosterone in the PRS+PB group (“± 476”) appears erroneous. Please amend.

Page 7, line 234. Please verify whether “effect” and “trail” are correct (likely “affect” and “trial”).

Answers: All corrected - thank you for your diligence

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

The authors have improved the manuscript.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have responded thoroughly to all my concerns and have corrected the grammatical errors. I would suggest writing the second Latin name of plant species in lowercase (e.g., Curcuma longa instead of Curcuma Longa). Apart from this, I have no further comments.