Preoperative Phase Angle as a Marker of Nutritional and Functional Vulnerability in Patients Undergoing Colorectal Cancer Surgery: A Prospective Observational Cohort Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Major Comments

1. The authors state that "no formal a priori sample size calculation was performed" (Section 2.2). This is a significant methodological limitation. The study reports several subgroup analyses and secondary outcomes (e.g., readmission-related complications with only 35 events), yet the sample size appears insufficient for many of these analyses. The authors should either:

• Provide a post-hoc power analysis to demonstrate which comparisons are adequately powered
• Acknowledge this limitation more prominently in the discussion
• Consider whether underpowered analyses (particularly for readmission outcomes with wide confidence intervals) should be presented as exploratory findings

2. The authors deliberately report only unadjusted odds ratios (Section 2.6) and justify this by stating that "the primary aim of the study was to evaluate the discriminative and clinical performance of phase angle... rather than causal multivariable modelling." While this rationale is acknowledged, the lack of any adjusted analyses substantially limits the interpretability of the findings. Phase angle is known to be influenced by age, sex, body mass index, and hydration status. Without adjusting for these confounders, the observed associations may be confounded. For example, the association between phase angle <4.7° and readmission-related complications (OR 2.50, 95% CI 1.19-5.24) may simply reflect that older patients have both lower phase angle and higher complication rates.

3. The use of "any postoperative complication" as a composite outcome is problematic. As the authors acknowledge, this composite includes heterogeneous events. However, they then present phase angle's poor discrimination for this outcome as a key finding. This is somewhat tautological, a marker of nutritional/functional vulnerability would not be expected to predict all complications equally.
4. The authors used complete-case analysis and excluded patients with missing data (Section 2.6). This approach can introduce bias if data are not missing completely at random. The authors report that readmission analyses included only 203 patients (98% of the cohort) and hospital stay analyses included only 199 patients (96%). While these proportions are relatively small, the reasons for missing data are not described.
5. The proposed thresholds of 5.0° for malnutrition and 4.7° for functional vulnerability are presented with confidence, yet the authors appropriately caution that phase angle values depend on device, population, and measurement conditions (Section 4). This creates a tension: the authors want to provide clinically actionable cut-offs while acknowledging they may not be generalizable.
6. Patients were excluded if they had "limitations in oral intake or any clinical condition contraindicating use of the oral route" (Section 2.2). This is a potentially important source of selection bias, as these patients may represent a particularly vulnerable group in whom phase angle would have the greatest utility.
7. The overall postoperative complication rate was 46.9% (97/207), which appears higher than rates typically reported in enhanced recovery programmes for colorectal surgery. This raises questions about the representativeness of the cohort or the definitions used.
8. In Table 3, the AUC for GLIM-defined malnutrition (0.728) and SPPB <10 (0.785) are described as "good discrimination," yet the 95% confidence intervals suggest moderate precision. The terminology used to describe AUC values should be consistent with established conventions (e.g., 0.70-0.80 = acceptable, 0.80-0.90 = excellent).
9. The discussion appropriately highlights the dissociation between phase angle's association with nutritional/functional outcomes versus overall complications. However, the interpretation that phase angle "does not show sufficient capacity to discriminate the occurrence of any postoperative complication" should be tempered by the recognition that other well-validated risk scores (e.g., ASA, POSSUM) also have limited discriminative ability for any-complication composites.
10. The use of ROC curves to determine optimal cut-offs is appropriate, but the authors do not specify which criterion was used for selecting the optimal threshold (e.g., Youden's index, highest sensitivity-specificity balance).
11. The authors suggest that low phase angle should "activate closer assessment of nutritional and functional reserve" (Section 4). However, this is already standard practice in enhanced recovery programmes. The incremental value of adding phase angle to existing nutritional and functional assessments is not clearly demonstrated.

Comments on the Quality of English Language

The manuscript requires a thorough proofreading pass to correct typographical errors (duplicated words, medical misspellings like 'dynamenia'), fix duplicated section headings, and refine a few awkward word choices.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This study investigates the role of preoperative phase angle as a marker of nutritional and functional vulnerability in patients undergoing colorectal cancer surgery. The authors conducted a prospective observational cohort study with 207 patients, assessing phase angle via bioelectrical impedance and correlating it with GLIM-defined malnutrition, handgrip strength, SPPB, and postoperative outcomes. The key contribution is the demonstration that phase angle is more strongly associated with nutritional and functional vulnerability than with overall postoperative morbidity, suggesting its utility in prehabilitation and risk stratification. However, some points merit further consideration:

Major concerns

1.The manuscript clearly addresses whether baseline phase angle is more strongly associated with nutritional/functional vulnerability than with postoperative morbidity. However, the rationale for focusing on composite morbidity outcomes could be better explained, given their heterogeneity.

2.While the methods are generally well described, the manuscript should clarify whether hydration status was standardized before BIA measurement, as this can significantly affect phase angle values. Additionally, the absence of a formal sample size calculation should be acknowledged more explicitly as a limitation.

3.The reliance on unadjusted odds ratios and ROC analyses is appropriate for exploratory purposes, but the lack of multivariable adjustment may limit causal interpretation. The authors should emphasize this limitation more strongly in the discussion.

4.ROC curves are informative, but confidence intervals should be displayed to enhance interpretability. Figure labeling could also be improved for clarity.

5.The discussion could more explicitly highlight how phase angle measurement might be integrated into routine preoperative assessment alongside GLIM and functional tests, and what thresholds should guide clinical decision-making.

Minor concerns

1.The English is generally fine, but some sentences are overly complex. Simplifying phrasing would improve readability (e.g., breaking long multi-clause sentences into shorter units).

2.The references are appropriate and relevant, but inclusion of more recent meta-analyses on phase angle and functional outcomes in oncology would strengthen the background.

3.Tables are clear and well-presented, but the supplementary flow diagram should be referenced more explicitly in the main text.

4. Ensure consistent use of terms such as “nutritional risk” versus “malnutrition” to avoid confusion.

Comments on the Quality of English Language

The manuscript is generally well written and understandable. However, several sentences are overly complex and could be simplified for clarity. Minor grammar refinements and smoother phrasing would improve readability. For example, long multi‑clause sentences in the introduction could be broken into shorter units to enhance flow. Overall, the English is adequate for publication but would benefit from light editing to ensure precision and ease of reading.

Author Response

Dear Reviewer 2,
We sincerely thank you for your constructive assessment and for recognising the contribution of our study. Your comments helped us clarify the rationale for the clinical outcomes, strengthen the analysis and limitations, improve the presentation of the ROC results, and define a more cautious role for PhA within preoperative assessment. We have addressed all comments point by point below.

●        General comment. This study investigates the role of preoperative phase angle as a marker of nutritional and functional vulnerability in patients undergoing colorectal cancer surgery. The authors conducted a prospective observational cohort study with 207 patients, assessing phase angle via bioelectrical impedance and correlating it with GLIM-defined malnutrition, handgrip strength, SPPB, and postoperative outcomes. The key contribution is the demonstration that phase angle is more strongly associated with nutritional and functional vulnerability than with overall postoperative morbidity, suggesting its utility in prehabilitation and risk stratification. However, some points merit further consideration.

Response: We appreciate this accurate summary of the study and its intended contribution. In response to the concerns below, we have refined the manuscript so that PhA is presented as a complementary, outcome-specific marker of nutritional and functional vulnerability rather than as a stand-alone predictor of postoperative morbidity or a replacement for established assessments.

Location in the revised manuscript: Abstract; Introduction; Discussion; Conclusions.

“PhA may therefore serve as a complementary marker that prompts comprehensive nutritional and functional assessment alongside GLIM, body composition, handgrip strength and SPPB.”

Major Concerns

●        Comment 1. The manuscript clearly addresses whether baseline phase angle is more strongly associated with nutritional/functional vulnerability than with postoperative morbidity. However, the rationale for focusing on composite morbidity outcomes could be better explained, given their heterogeneity.

Response: We agree. The Introduction now explains why the clinically common any-postoperative-complication composite was retained and explicitly states that weaker discrimination was expected because the composite combines events with different mechanisms and severity. The outcome has been classified as secondary and exploratory throughout. Methods clarify that it reflects routine surveillance rather than a homogeneous pathophysiological endpoint, and non-diarrhoeal complications are analysed separately. The Discussion interprets the result as an outcome-specific limitation rather than a general failure of PhA.

Location in the revised manuscript: Introduction, penultimate and final paragraphs; Section 2.3; Sections 3.2–3.4; Discussion, opening two paragraphs and limitations; Conclusions.

“The present study retained the clinically common ‘any postoperative complication’ composite as a secondary exploratory endpoint to examine whether PhA discriminates broad morbidity. Because the composite combines events with different mechanisms and severity, a weaker association was expected than for nutritional and functional outcomes.”

“This inclusive composite captured at least one recorded event of any listed type or severity, including diarrhoea; it was retained to reflect routine surveillance rather than a homogeneous pathophysiological endpoint.”

●        Comment 2. While the methods are generally well described, the manuscript should clarify whether hydration status was standardized before BIA measurement, as this can significantly affect phase angle values. Additionally, the absence of a formal sample size calculation should be acknowledged more explicitly as a limitation.

Response: We agree. Measurements were standardised with the participant supine after 5 minutes of rest and with a standard tetrapolar placement. However, the available records did not document fasting, recent exercise or fluid-intake restrictions. We therefore state explicitly that hydration was not standardised beyond the resting protocol. Extracellular water and hydration percentage were measured, and extracellular water was included in an exploratory sensitivity analysis. Hydration-related confounding and collinearity are now discussed. The absence of an a priori sample-size calculation is also described in Methods and highlighted in the limitations; sparse-event analyses are labelled exploratory and interpreted using event counts and 95% confidence intervals.

Location in the revised manuscript: Section 2.2, sample-size paragraph; Section 2.3, BIA protocol; Section 2.6, final paragraph; Discussion, hydration and limitations paragraphs.

“The records available for this analysis did not document fasting, recent exercise or fluid-intake restrictions; hydration was therefore not standardised beyond the resting measurement protocol.”

“Limitations include the single-centre design, a fixed sample recruited over a predefined period and the absence of an a priori sample-size calculation.”

●        Comment 3. The reliance on unadjusted odds ratios and ROC analyses is appropriate for exploratory purposes, but the lack of multivariable adjustment may limit causal interpretation. The authors should emphasize this limitation more strongly in the discussion.

Response: We agree and have gone beyond acknowledging the limitation by adding exploratory multivariable logistic regression models adjusted for age, sex and body mass index. A hydration-focused sensitivity model additionally included extracellular water percentage. Revised Table 4 presents both unadjusted and adjusted odds ratios, and the Results report the adjusted associations. Because sparse events constrained model dimensionality, no automated selection was used and other clinical factors were not modelled simultaneously. Methods and Discussion now state explicitly that residual confounding remains and that the estimates are associative, not causal.

Location in the revised manuscript: Sections 2.5–2.6; Section 3.3; revised Table 4; Discussion, hydration and limitations paragraphs; Conclusions.

“Exploratory multivariable logistic regression models then estimated adjusted odds ratios for PhA ≤4.7°, including age, sex and body mass index as covariates.”

“The models did not account simultaneously for all clinical factors; residual confounding remains, and the estimates are associative rather than causal.”

●        Comment 4. ROC curves are informative, but confidence intervals should be displayed to enhance interpretability. Figure labeling could also be improved for clarity.

Response: We agree. Figure 2 has been revised so that the curves and outcomes are identified more clearly and the legend displays the AUC with its nonparametric DeLong 95% confidence interval for each selected outcome. All outcome-specific AUC confidence intervals are also reported in revised Table 3. The caption and abbreviation line now define ROC, AUC, GLIM, SPPB and the handgrip 10th-percentile label.

Location in the revised manuscript: Section 3.2; revised Figure 2 and its legend; revised Table 3.

“ROC curves for selected outcomes are shown in Figure 2. The 95% confidence intervals for each AUC are displayed in the figure legend and reported for all outcomes in Table 3.”

“AUC estimates and nonparametric DeLong 95% confidence intervals are displayed in the legend. p10: 10th percentile.”

●        Comment 5. The discussion could more explicitly highlight how phase angle measurement might be integrated into routine preoperative assessment alongside GLIM and functional tests, and what thresholds should guide clinical decision-making.

Response: We agree and have clarified the proposed clinical role. PhA is now presented as a rapid complementary marker that may prompt or reinforce completion of GLIM assessment, body-composition evaluation, handgrip dynamometry and SPPB when these are not yet available. It should not replace those assessments or create a new independent ERAS standard. The cohort-derived values of 5.0° for GLIM-defined malnutrition and 4.7° for functional vulnerability are described as candidate screening thresholds selected by the Youden index, not as universal treatment or stand-alone action thresholds. External validation and formal incremental-value assessment are required before routine implementation.

Location in the revised manuscript: Section 3.3; Discussion, threshold, integration and ERAS paragraphs; Conclusions.

“PhA should complement rather than replace standard nutritional and functional assessments.”

“Within ERAS pathways, low PhA should prompt or reinforce the nutritional and functional assessments that are already recommended, not replace them or define a new independent standard of care.”

“Our values should therefore be described as cohort-derived candidate screening thresholds, not treatment thresholds.”

Minor Concerns

●        Comment 1. The English is generally fine, but some sentences are overly complex. Simplifying phrasing would improve readability (e.g., breaking long multi-clause sentences into shorter units).

Response: We agree. The manuscript has undergone a complete language edit. Long multi-clause sentences, particularly in the Introduction and Discussion, were divided or restructured; redundant wording was removed; and transitions were simplified. Grammar, punctuation and academic phrasing were also standardised while preserving the scientific meaning.

Location in the revised manuscript: Throughout the manuscript, particularly the Introduction and Discussion.

“Colorectal cancer surgery remains associated with morbidity despite minimally invasive techniques, anaesthetic optimisation and Enhanced Recovery After Surgery (ERAS) programmes. Outcomes also depend on preoperative physiological reserve.”

●        Comment 2. The references are appropriate and relevant, but inclusion of more recent meta-analyses on phase angle and functional outcomes in oncology would strengthen the background.

Response: We agree. The Introduction and Discussion now synthesise recent oncology evidence, including the 2023 systematic review and meta-analysis by Victoria-Montesinos et al. on PhA, handgrip strength and nutritional indicators; the 2024 study by Härter et al. on standardised PhA, function and muscle mass in surgical cancer patients; the 2024 systematic review by Prete et al.; and the 2025 cancer meta-analysis by Kong et al. These sources are cited within references 13–16 of the revised manuscript.

Location in the revised manuscript: Introduction, PhA evidence paragraph; Discussion, evidence-synthesis paragraph; references 13–16.

“Evidence from oncology scoping reviews, prospective studies, systematic reviews and meta-analyses links lower PhA to lean mass depletion, malnutrition, reduced handgrip strength, impaired physical function and poorer prognosis [12–16].”

●        Comment 3. Tables are clear and well-presented, but the supplementary flow diagram should be referenced more explicitly in the main text.

Response: This has been corrected. Figure S1 is now cited explicitly in the first paragraph of Results, where participant flow, exclusions and outcome-specific data availability are reported. The Supplementary Materials statement also identifies Figure S1 by title.

Location in the revised manuscript: Section 3.1, first paragraph; Supplementary Materials statement.

“Participant flow, exclusions and outcome-specific data availability are shown in Figure S1.”

●        Comment 4. Ensure consistent use of terms such as “nutritional risk” versus “malnutrition” to avoid confusion.

Response: We agree. The terminology has been standardised and an explicit definition has been added. ‘GLIM-defined malnutrition’ is reserved for the diagnostic outcome based on GLIM criteria. ‘Baseline nutritional vulnerability’ is used only for the descriptive combined category of malnutrition or nutritional risk. Table titles, Results and the abstract were reviewed to maintain this distinction.

Location in the revised manuscript: Section 2.3, first paragraph; Sections 3.1–3.3; Tables 1–4; Abstract.

“For descriptive comparisons, ‘baseline nutritional vulnerability’ denotes the combined classification of malnutrition or nutritional risk; it is distinct from the specific outcome ‘GLIM-defined malnutrition’.”

Comments on the Quality of English Language

●        Comment 1. The manuscript is generally well written and understandable. However, several sentences are overly complex and could be simplified for clarity. Minor grammar refinements and smoother phrasing would improve readability. For example, long multi-clause sentences in the introduction could be broken into shorter units to enhance flow. Overall, the English is adequate for publication but would benefit from light editing to ensure precision and ease of reading.

Response: We agree and thank the reviewer. A thorough English-language proofreading pass has been completed. Complex sentences were simplified, grammar and punctuation were corrected, awkward phrasing was refined, duplicated wording and headings were removed, and terminology and abbreviations were standardised. Particular attention was given to the Introduction, Discussion, table titles, figure legends and statistical reporting.

Location in the revised manuscript: Throughout the revised manuscript.

“The complete manuscript has been edited for precision, clarity and ease of reading.”

We again thank the reviewer. We believe that the revisions have substantially strengthened the manuscript.
Sincerely,
The Authors

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript investigates the clinical performance of baseline phase angle as a marker of nutritional and functional vulnerability in patients with colorectal cancer scheduled for elective surgery within a multimodal rehabilitation pathway. The work is generally clear; however, it requires some refinement before it can be accepted for publication. Here are my main concerns about this manuscript.

Specific comments:

1. Revise the manuscript carefully to correct typos, occasional grammatical errors, and awkward phrasing.
2. Abstract: Clarify which outcomes were primary vs. secondary.
3. The introduction is lengthy. Condense the main ideas.
4. Lines 122–132: End the introduction with a clear statement on the hypothesis.
5. Line 155: “No formal a priori 156 sample size calculation was performed”; explain the reason.
6. Lines 192–198: The GLIM criteria are repeated. Rewrite to reduce repetition and refer to the reference.
7. Lines 222–239: Were postoperative outcomes collected by blinded assessors to reduce bias?
8. Add a schematic diagram to summarize the experimental design.
9. Lines 273–282: Distinguish between exploratory and confirmatory analyses.
10. Lines 284–295: Discuss the impact of missing data.
11. The results are overly detailed and repetitive. Each result is transcribed in full. Rewrite in a concise manner.

11. Lines 369–371: Quantify the separation.

1. Table 4: Consider grouping the outcomes by type.
2. Lines 474–495: This part repeats results at length. Summarize the findings and avoid repetition.
3. Lines 523–533: Avoid statements like “This distinction is clinically reasonable”.
4. Expand on the mechanisms that link low PhA to increased surgical risk or poor outcomes in colorectal cancer patients.
5. Do interventions based on low preoperative PhA improve outcomes?
6. Are changes in PhA perioperatively or postoperatively associated with outcomes?
7. Expand on the thresholds justification and how they compare to previously reported cut-offs.
8. Discuss the effects of hydration status, inflammation, comorbidities, and medication use on PhA or the outcomes.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Accept in present form

Reviewer 3 Report

Comments and Suggestions for Authors

The revised manuscript demonstrates significant improvement following the revisions. The authors have thoughtfully addressed all of the suggested comments, enhancing the clarity and overall quality of the submission.