Novel Hypothesis on Phase Angle as a Candidate Marker Associated with Physiological Reserve in Geriatric Patients in Long-Term Care

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Overall:

This paper tackles a clinically important and genuinely under-studied population (institutionalized, very old (mean age 87.5), high-comorbidity long-term-care residents) in whom bioimpedance data are scarce, and it does so with a coherent narrative linking phase angle (PhA) to frailty, sarcopenia, functional status and quality of life. The paper is overall good and can make a good addition to the literature, although it needs revision. 

Major recommendations:

1. Needs to reposition the contribution relative to prior work. The hypothesis that phase angle indexes biological/physiological reserve has meaningful precedent, so the “first hypothesis” framing in the Highlights will likely draw challenge. Zanforlini 2019 [PMID 31028839] is a prospective study that already concludes PhA “may be indirect measures of functional reserve, with lower values being potential biomarkers of evolving frailty.” It is not currently cited and should be added. The authors can thus position their work as extending and testing this hypothesis in institutionalized long-term-care residents.

2. Align claim strength with the cross-sectional design. Physiological reserve seems to be, by definition, a dynamic property as it is the capacity to respond to a stressor. Thus a cross-sectional study can demonstrate that PhA correlates with reserve-related markers, but cannot establish that PhA measures the response capacity itself. Softening a few phrases (e.g., “provides the missing piece of the puzzle” to “provides supporting cross-sectional evidence for”) would make the conclusions robust. The authors’ planned longitudinal follow-up (mentioned in the Conclusion) is the right next step and could be foregrounded.

3.  Frame the sarcopenia models with an explicit limitation, and specify the EWGSOP2 operationalization. Because the EWGSOP2 sarcopenia label here uses BIA-estimated muscle mass, and PhA and BCM come off the same impedance signal, the combined PhA+BCM discrimination (AUC 0.893) seems optimistic due to the shared inputs. Grip strength is the one independent component, so the coupling is partial. Two proportionate fixes: (i) state this shared-measurement limitation explicitly, and (ii) specify the EWGSOP2 operationalization actually used, including the ASM/SMI cut-off, and whether confirmed sarcopenia required both low strength and low mass.

4. Reconcile the multimorbidity figures:

• Text states multimorbidity was “13.7%”, but 7/45 = 15.6%.
• Table 1 frailty split lists 7 (25.9%) in the frail group (n = 23); 7/23 = 30.4%.
• The frailty split totals 7 (Table 1) multimorbid patients while the sarcopenia split totals 6 (5+1). Reconcile so both sum to the same N.
• The Discussion calls multimorbidity “a predominant finding”, but at ~15% this seems inconsistent. 

Minor / editorial suggestions:

1. Length of stay: text reports 7.3 ± 4.8 yr, but Table 1’s overall value is 7.38 ± 5.94.  reconcile.
2. Light proofreading: “stadistics” (Author contributions), “quality of fife” (Table 2 caption), “This: research” (Funding).
3. Table 2 contains several untranslated Spanish headings (e.g., "Demográficos," "Somatometria," "Bioimpedancia")
4. Calling inline figure and equation items "Appendix 1/2" is confusing, and a reader can mistake them for missing supplementary material

Author Response

We thank both reviewers for their careful and constructive evaluation. We have found the comments valuable and have revised the manuscript accordingly. Below we respond to each comment individually; reviewer comments are shown in italics and our responses follow, with the corresponding changes indicated by manuscript section.

Comment 1 — Reposition the contribution relative to prior work; cite Zanforlini 2019.

Reviewer: The “first hypothesis” framing will likely draw challenge; Zanforlini 2019 (PMID 31028839) already concludes PhA may be an indirect measure of functional reserve and is not cited. Reposition as extending and testing this hypothesis in institutionalized long-term-care residents.

Response: We agree and have made all three changes. (i) The “first hypothesis” wording has been removed from the Highlights, which now read:

“Phase angle as a candidate biomarker of physiological reserve in long-term-care residents.”

(ii) We now cite Zanforlini et al. (2019) and (iii) have added a passage to the Discussion positioning our work as an extension and test of this hypothesis in a distinct population. We also clarify a conceptual point the reviewer’s framing raises: Zanforlini interpreted phase angle as an indirect measure of functional reserve in fit community-dwelling adults (only ~2% frail at entry), whereas we examine it as an expression of physiological reserve at the cellular level in institutionalized residents in whom reserve is markedly contracted. The added text reads:

“Zanforlini et al. (2019) proposed that phase angle, together with metabolic equivalents, may serve as an indirect measure of the functional reserve of the organism, with lower values behaving as potential biomarkers of evolving frailty. That study, however, was conducted in fit community-dwelling older adults … and used functional reserve as an interpretive label without defining or operationalizing the construct. Building on that observation, the present study reframes phase angle specifically as a quantifiable expression of physiological reserve at the cellular level, and examines it in institutionalized long-term-care residents … our shift is therefore one of analytical level and population rather than a claim of conceptual disjunction.”

Comment 2 — Align claim strength with the cross-sectional design.

Reviewer: Physiological reserve is a dynamic property; a cross-sectional study can show PhA correlates with reserve-related markers but not that it measures response capacity. Soften phrases (e.g., “provides the missing piece of the puzzle” → “provides supporting cross-sectional evidence for”); foreground the planned longitudinal follow-up.

Response: Agreed. This has been addressed as part of the coordinated revision (see the general note above and Reviewer 2, points 1–2). We have softened representational language throughout, replacing “represents/quantifies/measures” with “is associated with / may reflect.” The specific phrase flagged by the reviewer, “provides the missing piece of the puzzle,” has been removed from the Abstract, whose conclusion now reads:

“In this cross-sectional study, phase angle was associated with frailty, sarcopenia, functional status, and quality of life — conditions linked to diminished physiological reserve in long-term-care older adults. These findings are hypothesis-generating and support, rather than demonstrate, the proposal that phase angle may reflect physiological reserve at the cellular level; longitudinal validation is required.”

Comment 3 — Frame the sarcopenia models with an explicit limitation and specify the EWGSOP2 operationalization.

Reviewer: Because the EWGSOP2 label uses BIA-estimated muscle mass, and PhA and BCM come off the same impedance signal, the combined PhA+BCM AUC (0.893) seems optimistic; grip strength is the one independent component. (i) State the shared-measurement limitation; (ii) specify the ASM/SMI cut-off and whether confirmed sarcopenia required both low strength and low mass.

Response: We agree and have addressed both parts. (i) We now state the shared-measurement limitation explicitly and reframe the interpretation of the combined AUC, additionally reporting the low intercorrelation between the two predictors:

“PhA and BCM showed only weak intercorrelation (Spearman ρ = 0.40; VIF ≈ 1.2), indicating negligible collinearity and non-redundant contributions. Nevertheless, both predictors and the BIA-estimated muscle-mass criterion of the EWGSOP2 label derive from the same bioelectrical signal; the combined-model AUC (0.893) should therefore be regarded as an apparent, internally optimistic estimate reflecting a shared measurement source, only partially attenuated by the impedance-independent handgrip component. Given the sample size (16 non-sarcopenic cases) and the absence of internal validation, this figure requires confirmation in larger cohorts.”

(ii) The Methods now fully specify the EWGSOP2 operationalization:

“Sarcopenia was diagnosed according to the EWGSOP2 algorithm. Muscle strength was measured with a Jamar® hydraulic dynamometer, with reduced strength defined as <27 kg in men and <16 kg in women. Skeletal muscle mass was estimated by bioelectrical impedance analysis, with low muscle mass defined as a skeletal muscle mass index (SMI) <7.0 kg/m² in men and <6.0 kg/m² in women. Confirmed sarcopenia required both reduced strength and low muscle mass; the 29 participants classified as sarcopenic met both criteria.”

 

Comment 4 — Reconcile the multimorbidity figures.

Reviewer: Text says 13.7%; 7/45 = 15.6%. Table 1 frail group lists 7 (25.9%); 7/23 = 30.4%. Frailty split totals 7 multimorbid; sarcopenia split totals 6 (5+1). Discussion calls multimorbidity “a predominant finding” at ~15%.

Response: We thank the reviewer for identifying these discrepancies. We returned to the source dataset and recomputed the figures. The correct number of participants with multimorbidity (≥5 diagnoses) is N = 6 (13.3%). The text and Table 1 have been corrected accordingly, and both stratifications now sum to this total (frailty split: 6 + 0; sarcopenia split: 5 + 1). In the Discussion, multimorbidity is no longer described as “a predominant finding”; we now clarify that, whereas multimorbidity in the strict sense (≥5 diseases) affected 13.3% of the sample, a burden of three or more chronic conditions was near-universal, as is typical of long-term-care residents. The revised passage reads:

“Multimorbidity … was a clinically relevant finding: although multimorbidity in the strict sense (≥5 diseases) affected 13.3% of our sample, a burden of three or more chronic conditions was near-universal, as is typical of long-term-care residents and implies a greater risk than in community-dwelling populations.”

Reviewer 1 — Minor / editorial

m1. Length of stay. Text reports 7.3 ± 4.8 yr vs. Table 1’s 7.38 ± 5.94.

m2. Typographical errors. Corrected: “stadistics” → “statistics” (Author contributions); “quality of fife” → “quality of life” (Table 2 caption); “This: research” → “This research” (Funding).

m3. Untranslated Table 2 headings. The Spanish headings (“Demográficos,” “Somatometría,” “Bioimpedancia”) have been translated (“Demographics,” “Anthropometry,” “Bioimpedance”).

m4. “Appendix 1/2” naming. The inline figure and equation previously labelled “Appendix 1/2” have been renamed as “Figure 2” and “Figure 4” to avoid the impression of missing supplementary material.

Reviewer 2 Report

Comments and Suggestions for Authors

I would like to congratulate the authors for addressing an important and clinically relevant topic. The concept of identifying an objective marker of physiological reserve in older adults is highly interesting, and the manuscript presents an original hypothesis with potential implications for geriatric assessment. The study is generally well organized and uses recognized geriatric assessment tools. However, several issues should be addressed before the manuscript can be considered for publication.

Major Comments

1. The conclusions are stronger than the evidence supports.

The study is cross-sectional and demonstrates associations between phase angle and frailty, sarcopenia, functional status, and quality of life. However, throughout the manuscript, particularly in the Discussion and Conclusions, the authors frequently imply that phase angle represents or quantifies physiological reserve. This conclusion cannot be established from the present study design.

2. The distinction between association and causality should be emphasized.

Because all variables were measured simultaneously, it is impossible to determine whether lower phase angle reflects reduced physiological reserve, whether frailty and sarcopenia lead to lower phase angle, or whether both are consequences of underlying biological aging. The Discussion should explicitly acknowledge this limitation.

3. The concept of physiological reserve requires a more balanced discussion.

Physiological reserve is a complex multidimensional construct involving cardiovascular, pulmonary, neuromuscular, endocrine, metabolic, immune, and cognitive systems.The manuscript currently assumes that cellular membrane integrity measured by phase angle adequately represents this multidimensional concept.The authors should better justify this conceptual framework and acknowledge alternative interpretations.

4. Sample size and statistical power should be discussed more thoroughly.

Only 45 participants were included, yet multiple regression analyses, ROC analyses, and numerous statistical comparisons were performed.Please discuss the potential risk of overfitting and limited statistical power.

5. Multiple comparisons should be addressed.

A large number of bioimpedance variables were tested against several clinical outcomes.The authors should clarify whether any adjustment for multiple testing was considered or explain why no correction was applied.

6. Selection of variables in the multivariable models requires clarification.

The rationale for selecting variables in the regression models is insufficiently described.Please explain why certain covariates were included, whether clinically important confounders were considered, how collinearity was assessed.

7. Potential confounding variables should be discussed.

Several factors known to influence phase angle were not included in the analyses, such as nutritional status, inflammatory markers, hydration status, renal function, or cognitive impairment.Although not all variables may have been available, this limitation should be acknowledged.

8. External validity is limited.

The study was conducted in a single long-term care institution with a very old and highly dependent population. The applicability of the proposed cut-off values to community-dwelling older adults or other healthcare settings is uncertain and should be discussed.

9. ROC-derived cut-off values require cautious interpretation.

The reported cut-off values were derived from the same dataset in which performance was evaluated. Without external validation, these thresholds should be considered exploratory.

10. The Discussion should better reflect conflicting evidence.

Although many studies support the usefulness of phase angle, several recent publications have reported limited or inconsistent predictive performance.Including these studies would provide a more balanced discussion.

Minor Comments

11. The title may overstate the findings.

The current title suggests that phase angle is a quantitative parameter of physiological reserve. Consider wording that more clearly reflects the hypothesis-driven nature of the study.

12. The Abstract should be more cautious.

The final sentence should emphasize that the findings support a hypothesis rather than demonstrate a new biomarker.

13. Language editing

The manuscript is generally well written but would benefit from professional English language editing to improve readability and remove occasional grammatical inconsistencies.

14. Future studies

The manuscript would benefit from a short paragraph outlining the types of longitudinal or prospective studies required to validate phase angle as a biomarker of physiological reserve.

Author Response

We thank both reviewers for their careful and constructive evaluation. We have found the comments valuable and have revised the manuscript accordingly. Below we respond to each comment individually; reviewer comments are shown in italics and our responses follow, with the corresponding changes indicated by manuscript section.

Comment 1 — Conclusions stronger than the evidence supports.

Reviewer: The manuscript frequently implies that phase angle represents or quantifies physiological reserve, which the cross-sectional design cannot establish.

Response: We agree. As described in the general note above, we have carried out a manuscript-wide revision that reframes the central claim from representation/quantification to association and hypothesis. Representational verbs (“is / represents / constitutes / quantifies / a parameter of”) have been replaced with association- and hypothesis-level language (“is associated with / may reflect / candidate marker / we hypothesize that”) in the title, abstract, discussion and conclusions. See our responses to points 2, 3, 11 and 12 for the specific edits.

Comment 2 — Emphasize the distinction between association and causality.

Reviewer: With simultaneous measurement it is impossible to determine whether lower PhA reflects reduced reserve, whether frailty/sarcopenia lower PhA, or whether both are consequences of biological aging. The Discussion should acknowledge this.

Response: Agreed. We have added an explicit statement to the Limitations:

“Its cross-sectional design precludes any causal or directional inference: a lower phase angle may reflect reduced physiological reserve, frailty and sarcopenia may themselves lower phase angle, or all three may be parallel consequences of underlying biological aging. These possibilities cannot be distinguished with simultaneously measured variables.”

Comment 3 — More balanced discussion of the physiological-reserve construct.

Reviewer: Physiological reserve is multidimensional (cardiovascular, pulmonary, neuromuscular, endocrine, metabolic, immune, cognitive). The manuscript assumes membrane integrity adequately represents it. Justify the framework and acknowledge alternatives.

Response: Agreed. We have added a dedicated paragraph to the Discussion:

“Physiological reserve is a multidimensional construct that integrates cardiovascular, pulmonary, neuromuscular, endocrine, metabolic, immune and cognitive systems, and no single parameter can represent it in full. Phase angle captures one integrative axis of this construct — the integrity of cell membranes and the mass of metabolically active tissue — that plausibly underlies the performance of these systems, but it does not directly quantify organ-specific reserve. Alternative interpretations must therefore be kept in view: a low phase angle may also reflect shifts in the ECW/ICW ratio, malnutrition or systemic inflammation rather than reserve as such. Accordingly, we position phase angle as a complementary, integrative cellular marker, to be interpreted alongside — not in place of — established multidimensional geriatric assessment.”

Comment 4 — Sample size and statistical power.

Reviewer: Only 45 participants, yet multiple regression, ROC and many comparisons were performed. Discuss overfitting and limited power.

Response: Agreed. The Limitations now address this explicitly:

“With 45 participants and several predictors per model, the analyses are exposed to overfitting and limited statistical power; the number of events per variable fell below recommended thresholds in some models, so the reported estimates should be regarded as exploratory and potentially optimistic.”

Comment 5 — Multiple comparisons.

Reviewer: Many bioimpedance variables were tested against several outcomes. Clarify whether adjustment for multiple testing was considered, or explain why none was applied.

Response: We have clarified our position in the Limitations: no formal correction was applied because the study is exploratory and hypothesis-generating, and p-values are interpreted descriptively rather than confirmatorily:

“Multiple bioimpedance parameters were tested against several clinical outcomes without formal correction for multiple comparisons; because the study is exploratory and hypothesis-generating, p-values are interpreted descriptively, and some associations may represent type I error and require independent confirmation.”

Comment 6 — Variable selection in the multivariable models.

Reviewer: The rationale for selecting variables is insufficiently described. Explain why covariates were included, whether confounders were considered, and how collinearity was assessed.

Response: We have expanded the Methods to describe the covariate-selection rationale and the collinearity assessment. Covariates were chosen a priori on clinical and pathophysiological grounds rather than by automated stepwise procedures, and the model was restricted to three clinically prioritized predictors to limit overfitting. The added text reads:

“Covariates were selected a priori on clinical and pathophysiological grounds rather than by automated stepwise procedures, and the model was restricted to three clinically prioritized predictors to limit overfitting. Age was included because it is consistently associated with phase angle in the literature and is central to our hypothesis, whereas functional status (Barthel) and frailty (FRAIL) were included as clinically established expressions of diminished physiological reserve. Collinearity among covariates was assessed with Spearman and Kendall rank correlations (n = 45): age was essentially uncorrelated with the other covariates (ρ ≤ 0.10), while frailty and functional status showed a moderate inverse correlation (Spearman ρ = −0.54, p = 0.001; approximate VIF ≈ 1.4), consistent with their shared physical-performance component. The two constructs are not redundant, however, as the FRAIL scale additionally captures comorbidity, exhaustion and weight loss that the Barthel index does not, and both retained independent, statistically significant associations with phase angle in the adjusted model (Table 3). No multicollinearity of concern was detected.”

Comment 7 — Potential confounding variables.

Reviewer: Factors known to influence PhA (nutritional status, inflammatory markers, hydration, renal function, cognition) were not included. Acknowledge this limitation.

Response: Agreed. We now acknowledge this, distinguishing partially captured from unmeasured factors:

“Several determinants known to influence phase angle were not formally modelled as confounders. Some were partially captured but not entered into the models — nutritional status through anthropometric indicators (BMI, calf and arm circumference) and hydration through BIA-derived water compartments (ECW/ICW) — whereas systemic inflammation, renal function, and formal cognitive assessment were not evaluated. Although functional status, which partly reflects advanced cognitive decline, was included, no validated cognitive instrument (e.g., MMSE/MoCA) was applied; cognitive impairment therefore remains a potential residual confounder.”

Comment 8 — Limited external validity.

Reviewer: Single long-term-care institution, very old and highly dependent population; applicability of the cut-offs to community-dwelling older adults or other settings is uncertain.

Response: Agreed. The Limitations now state this:

“The study was conducted in a single long-term-care institution with a very old, highly dependent population and among the lowest phase-angle values reported in the literature; the proposed cut-offs may not generalize to community-dwelling older adults or other care settings.”

Comment 9 — ROC-derived cut-offs require cautious interpretation.

Reviewer: Cut-offs were derived from the same dataset in which performance was evaluated; without external validation they should be considered exploratory.

Response: Agreed. We now label all cut-off values as internally derived and exploratory in the Results, Discussion and Abstract:

“All cut-off values were derived and evaluated within the same dataset, without external validation, and should be interpreted as exploratory rather than diagnostic thresholds.”

Comment 10 — Reflect conflicting evidence.

Reviewer: Several recent publications report limited or inconsistent predictive performance of PhA; including them would balance the discussion.

Response: Agreed. We have added such evidence, drawing on a study already in our reference list:

“This usefulness, however, is not uniformly supported. While many studies endorse the clinical value of phase angle, recent work has reported limited or inconsistent predictive performance; Piglowska et al., for instance, found a limited predictive value of bioelectrical phase angle for the development of sarcopenia in older Europeans. Such discrepancies likely reflect differences in study populations, bioimpedance devices and measurement frequencies, and the absence of standardized, age- and comorbidity-adjusted cut-off values — reinforcing that phase angle should currently be regarded as a promising but not yet validated marker.”

Comment 11 — The title may overstate the findings.

Reviewer: The current title suggests PhA is a quantitative parameter of physiological reserve; consider wording that reflects the hypothesis-driven nature of the study.

Response: Agreed. We have revised the title to a hypothesis-framed wording:

“Novel hypothesis on the role of phase angle as a candidate marker associated with physiological reserve in geriatric patients in long-term care.”

Comment 12 — The Abstract should be more cautious.

Reviewer: The final sentence should emphasize that the findings support a hypothesis rather than demonstrate a new biomarker.

Response: Agreed. The final sentence of the Abstract has been revised (see the quotation under Reviewer 1, point 2) so that it states the findings are hypothesis-generating and “support, rather than demonstrate” the proposal, with longitudinal validation required.

Comment 13 — Language editing.

Reviewer: The manuscript would benefit from professional English language editing.

Response: The manuscript has been carefully re-read and edited by the authors.

Comment 14 — Future studies.

Reviewer: Add a short paragraph outlining the longitudinal/prospective studies required to validate PhA as a biomarker of physiological reserve.

Response: Agreed. We have added such a paragraph to the Discussion:

“Validating phase angle as such a marker will require prospective, longitudinal cohort studies that measure phase angle repeatedly and relate it to hard clinical outcomes; incident frailty and disability, hospitalization, and mortality, thereby establishing temporality, which a cross-sectional design cannot. External validation of the proposed cut-offs in independent long-term-care and community-dwelling cohorts is equally necessary, ideally with adjustment for nutritional status, inflammation, hydration and renal function to isolate the independent contribution of phase angle. Our group is currently pursuing this line through studies using bioimpedance devices adapted for immobile patients and the development of a geriatric database of age- and comorbidity-adjusted reference values.”

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Concerns have been addressed.