Optimal Cutoff Values of Site-Specific Phase Angle for Predicting Independence in Activities of Daily Living in Post-stroke Patients
Article information
Abstract
Background
Phase angle (PhA) is a potential indicator of nutritional status and functional outcomes. However, optimal regional PhA cutoff values for predicting activities of daily living (ADL) post-stroke are unclear. We aimed to determine these cutoffs and assess whether PhA adds prognostic value to established clinical models.
Methods
In this retrospective cohort study, stroke patients undergoing inpatient rehabilitation had body composition and PhA measured via multi-frequency bioelectrical impedance analysis. The primary outcome was ADL independence at discharge (motor Functional Independence Measure >78). Receiver operating characteristic (ROC) curves identified optimal PhA cutoffs; logistic regression assessed their predictive value, adjusting for confounders. Incremental discriminative ability was evaluated using area under the ROC curve (AUC) comparisons with DeLong’s test.
Results
Among 1,080 patients (median age, 75.6 years; 54.1% men), optimal PhA cutoffs were: whole-body (women 3.90°, men 4.60°), non-paretic upper limb (4.45°, 4.90°), and non-paretic lower limb (4.00°, 4.80°). Higher PhA values above these cutoffs were independently associated with ADL independence (all p<0.05). Adding whole-body PhA to the baseline model improved AUC from 0.937 to 0.970 (ΔAUC=0.033; p=0.011); significant gains were observed for limb PhA.
Conclusions
Optimal site-specific PhA cutoffs predict ADL independence at discharge in post-stroke patients. PhA provides significant incremental prognostic value beyond established clinical models, supporting its use in tailoring rehabilitation strategies.
INTRODUCTION
Phase angle (PhA), a parameter obtained from bioelectrical impedance analysis (BIA), has attracted growing interest as a prognostic indicator of clinical outcomes across diverse patient populations.1-5) This metric reflects cellular integrity, membrane function, and hydration status, offering valuable information about an individual's overall health and nutritional condition.3,4) Prior investigations have revealed correlations between PhA and mortality, hospital length of stay, and functional outcomes in critically ill patients, as well as those with cardiovascular diseases and malignancies.5,6) Furthermore, its utility has been demonstrated in assessing cognitive decline in dementia, predicting functional recovery in osteoporosis, and monitoring muscle quality after spinal cord injury.7-9) The underlying mechanisms linking PhA to clinical outcomes are thought to involve its ability to detect early changes in body composition, such as fluid shifts and loss of body cell mass, which may reflect malnutrition or disease progression.6,10) Considering the potential clinical implications of PhA as a non-invasive, bedside assessment tool, additional research is warranted to explore its utility in predicting outcomes and guiding interventions in specific patient populations, such as those undergoing stroke rehabilitation.
However, evidence regarding the optimal cutoff values for segmental PhA in predicting functional outcomes in stroke patients with hemiparesis remains limited.2,11-13) While a few studies have investigated the association between segmental PhA and motor function or activities of daily living (ADL) in stroke rehabilitation, the specific cutoff values that best discriminate between favorable and unfavorable outcomes have not been well established.2) The mechanisms linking segmental PhA to functional recovery in stroke patients may involve changes in body composition, fluid distribution, and cellular health in the affected limbs.2,13) Lower PhA in the paretic limbs could reflect muscle atrophy, impaired cell membrane integrity, and fluid shifts, which may hinder motor recovery and ADL performance.2,13) Identifying optimal cutoff values for segmental PhA could have important clinical implications, as they may help predict prognosis, guide rehabilitation strategies, and monitor the effectiveness of interventions in stroke patients.2)
While paretic side PhA reflects localized muscle changes post-stroke, non-paretic side and whole-body PhA provide valuable insights into overall nutritional status and compensatory mechanisms. Non-paretic limb PhA may indicate increased reliance on the unaffected side for functional activities, potentially predicting rehabilitation outcomes.2) Recent evidence suggests that phase angles of healthy upper and lower limbs show stronger associations with functional independence and recovery compared to whole body or paretic limb measurements.14) Whole-body PhA offers a comprehensive assessment of cellular health and nutritional status, which are crucial for recovery.1) Additionally, comparing paretic and non-paretic side PhA can reveal asymmetries in body composition and muscle quality, informing tailored rehabilitation strategies.13) Hence, investigating non-paretic side and whole-body PhA alongside paretic side measurements provides a more complete picture of a stroke patient's functional potential and recovery trajectory.
Therefore, identifying optimal cutoff values for site-specific phase angle for whole-body and non-paralytic sides for predicting favorable outcomes in post-stroke patients would be essential for predicting recovery and facilitating treatment in this patient group. The aim of this study was to determine the optimal cutoff values for ADL independence for whole-body and site-specific phase angles, including the non-paralytic limb, in a cohort of post-stroke patients.
MATERIALS AND METHODS
Participants and Setting
A retrospective cohort study was performed at a single institution with a 135-bed dedicated stroke rehabilitation unit. The study cohort comprised all stroke patients sequentially admitted and discharged over an 8-year period from 2016 to 2023. Exclusion criteria were impaired consciousness on admission, missing data, or lack of informed consent. For each study participant, monitoring and data acquisition were continued until hospital discharge.
A comprehensive, multidisciplinary post-acute rehabilitation program was delivered by an interprofessional healthcare team for up to 3 hours per day throughout the patient's inpatient stay. The convalescent rehabilitation was personalized based on each individual's unique functional abilities and deficits, integrating a broad spectrum of therapeutic modalities. These encompassed physical therapy, occupational therapy, speech-language pathology, and audiology services, in conjunction with nutritional support,15) oral management,16) and pharmacological management.17) The physical therapy component included various interventions, such as facilitation of paralyzed limbs, range of motion exercises, gait training, resistance exercise, and training in ADLs.18) This holistic strategy was designed to maximize the patient's recovery and functional outcomes during the convalescent period following their acute illness or injury.
Data Collection
Upon admission to the convalescent rehabilitation wards, the following demographic and clinical characteristics were documented for each patient: age (in years), biological sex, type of stroke, time from stroke onset to admission (in days), and history of prior stroke. Baseline pre-stroke functional independence was assessed using the modified Rankin Scale (mRS),19) while comorbidity burden was quantified using the Charlson Comorbidity Index (CCI).20) The severity of post-stroke hemiparesis was evaluated using the Brunnstrom Recovery Stages (BRS) for the upper extremities, hands/fingers, and lower extremities.21) Functional independence in ADL and cognition were measured using the motor and cognitive subscales of the Functional Independence Measure (FIM), respectively, as assessed by experienced physical and occupational therapists.22) Body composition parameters, including the skeletal muscle index (SMI), were obtained via BIA using the InBody S10 (InBody Co. Ltd., Tokyo, Japan) within 3 days of admission.17) Handgrip strength was also measured at admission by rehabilitation professionals.23) Visual estimation was used by nurses or registered dietitians to quantify the proportion of food and drink consumed relative to the amount provided. Average daily energy and protein intake during the first week were calculated by dividing the total amounts by the patient's body weight in kilograms. The amount of inpatient rehabilitation was captured as the mean number of daily therapy units delivered, with each 20-minute session counting as one unit based on health insurance regulations.24) Additional data collected included the number of oral medications prescribed at the time of admission and the duration of inpatient rehabilitation in days.25)
Phase Angle Assessment
PhA was assessed using the InBody S10, a multi-frequency BIA device designed for use in supine or seated positions. This portable analyzer enables accurate, non-invasive body composition and body water analysis in patients who are immobile or have amputated limbs. The InBody S10 utilizes direct segmental multi-frequency BIA technology, measuring impedance at six frequencies (1, 5, 50, 250, 500, and 1,000 kHz) across five body segments (right arm, left arm, trunk, right leg, and left leg). It calculates whole-body and segmental phase angles at 50 kHz, which serve as indicators of cellular health and membrane integrity.26) To ensure optimal conditions for PhA measurement, the BIA assessment was conducted within 3 days of admission.27) Patients were required to be well-hydrated, remain on bed rest for 4 hours after their last meal, and refrain from eating for 1 hour prior to the evaluation. Individuals with active fever, tremor, or acute illness were excluded from testing.27) Phase angle (φ) was calculated using the equation:
φ = arctangent (Xc/R) × (180/π),
where R represents the resistance and Xc represents the reactance of the right half of the body at 50 kHz.28) The resulting phase angle values were recorded for each patient, providing valuable insights into their cellular health and overall nutritional status. The following phase angle values were obtained for each patient: whole body, upper limb (non-paretic side), upper limb (paretic side), lower limb (non-paretic side), and lower limb (paretic side)
For patients with bilateral motor impairment, the extremity with the lower BRS (i.e., greater weakness) was classified as the paretic side, and the contralateral extremity as the non-paretic side; if BRS scores were identical, the limb judged clinically more dysfunctional by the attending physiatrist in consensus with the responsible physical and occupational therapists—based on manual-muscle testing, dexterity tasks, and standing-and-gait assessment—was designated paretic.
Outcomes
The primary outcome measure was ADL independence at discharge, defined as the motor domain of the FIM (FIM-motor) score greater than 78. The FIM-motor is a widely used, validated tool for assessing an individual's functional status in terms of motor ADL.22) It consists of 13 motor items, each scored on a 7-point scale (1=total assistance, 7=complete independence), with a total score ranging from 13 to 91. A FIM-motor score of >78 has been considered as a clinically meaningful cutoff for ADL independence in rehabilitation patients.29-32) This threshold has been validated in multiple studies, demonstrating its ability to discriminate between independent and dependent individuals in terms of motor function and self-care activities.
To ensure objectivity and minimize potential bias, the FIM was administered by rehabilitation professionals, including physical therapists, occupational therapists, and nurses, who were not involved in the study's data collection, evaluation, analysis, or interpretation of results.33) These clinicians were trained in the proper application of the FIM assessment. Following each assessment, the respective therapists and nurses engaged in discussions to reconcile any discrepancies and adjust their scores accordingly, thereby reducing inter-rater variability.
Sample Size Calculation
To determine the required sample size for comparing the proportion of patients achieving ADL independence (defined as FIM-motor >78) at discharge between two groups based on the median PhA-whole body score at admission, we performed a power analysis for a two-proportion test. Previous studies have reported that approximately 40% of patients achieve ADL independence at discharge, while 60% remain ADL dependent.34) Assuming that the group with higher PhA-whole body scores would have a 10% higher proportion of ADL independence compared to the lower score group, a minimum sample size of 393 patients per group (786 total) is required to detect this difference with 80% power at a significance level of 0.05.
Statistical Analysis
For parametric data, means±standard deviations were presented, whereas medians with interquartile ranges (IQR) were used for non-parametric data. Categorical variables were expressed as counts and percentages (%). Comparisons between the two groups were conducted using the appropriate statistical tests based on the nature of the variables. The t-test was employed for parametric data, while the Mann–Whitney U test was used for non-parametric data. Categorical variables were compared using the chi-square test.
Logistic regression analysis was performed to investigate whether site-specific PhA were independently associated with the ADL independence at discharge after adjusting for confounding factors.34) The selection of confounding factors was carefully conducted based on clinical expertise and findings from previous studies, considering that the primary outcome was ADL at discharge and the known sex differences in PhA.27) As a result, the following factors were adjusted for age, sex, days from onset to admission, stroke type, FIM-motor, FIM-cognition, pre-stroke mRS, CCI, BRS of the lower limb, and number of medications.35-39) The variance inflation factor (VIF) was employed to evaluate the presence of multicollinearity among the predictor variables. VIF values ranging from 1 to 10 were considered to signify a lack of substantial multicollinearity.
Receiver operating characteristic (ROC) curves were constructed to determine the optimal cutoff values of each site-specific PhA for predicting ADL independence at discharge, defined as a FIM-motor score of >78. The area under the ROC curve (AUC) was calculated to assess the discriminative ability of each PhA parameter. Youden's index, which maximizes the sum of sensitivity and specificity, was used to identify the optimal cutoff values for each PhA parameter in men and women separately.
To validate the optimal cutoff values for site-specific PhA derived from the ROC curves, patients were categorized into high and low PhA groups using these cutoffs. Logistic regression analysis was then performed to investigate whether high site-specific PhA were independently associated with ADL independence at discharge. All covariates included in the multivariable models were selected a priori based on clinical relevance and prior studies, without the use of automated selection procedures. The models were adjusted for the same set of confounding factors as in the previous logistic regression analyses, including age, sex, days from onset to admission, stroke type, FIM-motor, FIM-cognition, pre-stroke mRS, CCI, BRS of the lower limb, and number of medications. This approach allowed for a robust assessment of the predictive value of site-specific PhA, including whole body, non-paretic upper and lower limbs, in determining discharge ADL independence, while controlling for potential confounders. Multicollinearity was assessed by calculating the VIF for each covariate. A VIF value of <10 was considered acceptable, and the VIF range for each model was reported in the Results section. In addition, Spearman correlation coefficients were calculated among the five site-specific PhA variables to evaluate potential redundancy. The overall goodness-of-fit of the logistic regression models was evaluated using the Hosmer–Lemeshow test.
To quantify the incremental value of PhA we compared a clinical-baseline model (covariates only) with two sets of PhA-extended models: (1) five models that added each continuous site-specific PhA (whole-body, paretic and non-paretic upper limb, paretic and non-paretic lower limb) and (2) three models that added the ROC-derived high-PhA indicators (whole-body, non-paretic upper limb, non-paretic lower limb). For every model the AUC was computed and the change versus baseline (ΔAUC) was tested with DeLong’s method
Statistical significance was set at a p-value threshold of less than 0.05. The entirety of the statistical analyses were conducted using SPSS Statistics software version 21 (IBM, Armonk, NY, USA).
Ethics
The study protocol was reviewed and approved by the Institutional Review Board of the Kumamoto Rehabilitation Hospital (Approval ID: 191-220315). Due to the retrospective nature of the study, obtaining written informed consent from each participant was not feasible. As an alternative, an opt-out procedure was implemented, enabling participants to withdraw from the study at any time. The study was carried out in compliance with the ethical principles outlined in the 1964 Declaration of Helsinki and its subsequent amendments, as well as the Ethical Guidelines for Medical and Health Research Involving Human Subjects (Provisional Translation as of March 2015).
RESULTS
During the study period, 1,137 stroke patients were admitted to the rehabilitation wards. Of these, 46 patients were excluded due to incomplete data, and an additional 11 patients were excluded because of altered consciousness. The remaining 1,080 patients were included in the final analysis (Fig. 1).
Flowchart of participant screening, inclusion criteria, and follow-up.
Table 1 summarizes the baseline characteristics of the 1,080 stroke patients, stratified by sex. The median age was 75.6 years, with 54.1% being man. Cerebral infarction was most common (63.6%), followed by cerebral hemorrhage (29.7%), and subarachnoid hemorrhage (6.7%). Median onset-to-admission time was 14 days. At admission, patients showed moderate to severe ADL impairment, with a median total FIM score of 66. Women consistently had lower PhA values than men across all body segments, including whole body, upper limbs (healthy and paretic sides), and lower limbs (healthy and paretic sides).
Baseline comparison of patient characteristics between women and men
Table 2 presents the results of logistic regression analysis examining the association between site-specific PhA and ADL independence at discharge, adjusted for potential confounders. All site-specific PhA, including whole body, paretic and non-paretic upper limbs, and paretic and non-paretic lower limbs, were significantly associated with ADL independence at discharge. The strongest associations were observed for non-paretic upper limb PhA (odds ratio [OR]=1.609, 95% confidence interval [CI] 1.118–2.316) and non-paretic lower limb PhA (OR=1.586, 95% CI 1.186–2.122). All VIF values ranged from 1.14 to 3.38 across the models, indicating no multicollinearity. The Hosmer–Lemeshow test yielded p-values of 0.429, 0.314, and 0.486 for the models incorporating whole-body, non-paretic upper limb, and non-paretic lower limb PhA, respectively, demonstrating good model calibration. As shown in Supplementary Table S1, each continuous-PhA variable significantly improved model discrimination over the baseline configuration (ΔAUC range +0.004 to +0.013; all p<0.05), with whole-body PhA providing the greatest gain (ΔAUC=0.013, p=0.041).
Logistic regression analysis of ADL independence at discharge by site-specific phase angles
To further assess the interrelatedness of PhA variables, Spearman correlation coefficients were computed among the five site-specific PhA measurements. As presented in Supplementary Table S2, all pairwise correlations were statistically significant (p<0.001), but none exceeded ρ=0.85, suggesting acceptable independence among predictors.
ROC curves are shown in Figs. 2 and 3 to detect the optimal cutoff values of site-specific PhA for predicting ADL independence at discharge. The results revealed that whole-body PhA had the highest discriminative ability for predicting ADL independence at discharge in both women (AUC=0.809) and men (AUC=0.806). The optimal cutoff values for whole-body PhA were 3.90° for women (sensitivity 67.6%, specificity 81.8%) and 4.60° for men (sensitivity 65.3%, specificity 82.3%). In comparison of the upper and lower extremities on the non-paretic limb, non-paretic upper limb PhA showed good predictive performance (women, AUC=0.786; men, AUC=0.775), with optimal cutoffs of 4.45° for women and 4.90° for men. Similarly, non-paretic lower limb PhA demonstrated fair discriminative ability (women, AUC=0.763; men, AUC=0.788), with optimal cutoffs of 4.00° for women and 4.80° for men. These results indicate that the diagnostic accuracy of non-paralyzed side PhA for discharge ADL independence was higher for upper limb in women and lower limb in men. These ROC-derived thresholds were used to define binary variables representing “higher PhA” for subsequent regression analyses.
Receiver operating characteristic curves to identify the optimal phase angle cutoff for detecting ADL independence at discharge in woman (A) and man (B) using phase angle–whole body. (A) The optimal cutoff value is 3.90° in woman (sensitivity=0.676, specificity=0.818). (B) The optimal cutoff value is 4.60° in man (sensitivity=0.653, specificity=0.823). ADL, activities of daily living; AUC, area under the curve.
Receiver operating characteristic curves to identify the optimal phase angle cutoff for detecting ADL independence at discharge in woman (A) and man (B) using phase angle–non-paretic upper limb, and in woman (C) and man (D) using phase angle-non paretic lower limb. (A) The optimal cutoff value is 4.45° in woman (sensitivity=0.676, specificity=0.818). (B) The optimal cutoff value is 4.90° in man (sensitivity=0.653, specificity=0.823). (C) The optimal cutoff value is 4.00° in woman (sensitivity=0.659, specificity=0.734). (D) The optimal cutoff value is 4.85° in man (sensitivity=0.694, specificity=0.759). ADL, activities of daily living; AUC, area under the curve.
Table 3 presents the results of logistic regression analysis examining the association between higher site-specific PhA, based on the optimal cutoff values derived from the ROC curves, and ADL independence at discharge. Higher PhA-whole body (OR=2.106, 95% CI 1.286–3.449), higher PhA-non paretic upper limb (OR=2.164, 95% CI 1.283–3.649), and higher PhA-non paretic lower limb (OR=1.641, 95% CI 1.055–2.720) were all significantly associated with ADL independence at discharge, after adjusting for potential confounders. These findings validate the predictive value of the optimal cutoff values for site-specific PhA in determining functional outcomes in stroke patients. All VIF values for the models in Table 3 ranged from 1.12 to 3.25, indicating no multicollinearity. The Hosmer–Lemeshow test yielded p-values of 0.387, 0.412, and 0.471, confirming good model calibration.
Logistic regression analysis of ADL independence at discharge by higher site-specific phase angles
To further evaluate the incremental prognostic value of PhA beyond established clinical predictors, we compared the AUC of a baseline logistic regression model with that of models incorporating each high-PhA variable. All models were adjusted using the same set of covariates as in Table 3. As shown in Table 4, all three high-PhA indicators significantly improved model discrimination for ADL independence. The addition of whole-body PhA yielded the greatest improvement in AUC (0.970 vs. 0.937; ΔAUC=+0.033, p=0.011), followed by non-paretic upper limb (ΔAUC=+ 0.028, p=0.019) and non-paretic lower limb (ΔAUC=+0.025, p=0.027).
Incremental discrimination of the baseline clinical model after adding higher phase-angle indicators for predicting ADL independence at discharge
DISCUSSION
This study aimed to determine the optimal cutoff values of site-specific PhA for discharge ADL independence in post-stroke patients using a retrospective cohort study design. The key findings were as follows: (1) the optimal cutoff values of PhA for discharge ADL independence were 3.90 for women and 4.60 for men in whole-body PhA, 4.45° for women and 4.90° for men in non-paretic upper limb PhA, and 4.00° for women and 4.80° for men in non-paretic lower limb PhA; (2) the diagnostic accuracy of non-paralyzed side PhA for discharge ADL independence was higher for upper limb in women and lower limb in men; and (3) the addition of site-specific PhA to a clinically adjusted multivariable logistic regression model significantly improved the model’s discriminative performance, with whole-body PhA yielding the greatest increase in AUC.
The optimal cutoff values of site-specific PhA for discharge ADL independence were reported. This novel finding highlights the significance and utility of regional PhA measurements in predicting functional outcomes in post-stroke patients who exhibited hemiplegia. The whole-body PhA reflects overall cell membrane integrity and nutritional status, while non-paretic limb PhA provides insights into localized muscle quality and strength, which are crucial determinants of ADL performance.2) Higher PhA indicates better cell membrane function and muscle quality,12) facilitating improved mobility and self-care abilities. Notably, the diagnostic accuracy varied between sex, with non-paretic upper limb PhA being more predictive in women and lower limb PhA in men, potentially due to differences in muscle distribution and functional requirements. These cutoffs could guide rehabilitation strategies, emphasizing targeted exercise and training for limbs below the threshold to enhance muscle quality and strength. Moreover, monitoring regional PhA changes may enable early detection of functional decline, prompting timely interventions.40) Incorporating site-specific PhA assessment using specific cutoffs into routine stroke care could optimize functional recovery and discharge planning, ultimately improving patient outcomes and reducing healthcare costs associated with disability.
The diagnostic accuracy of non-paralyzed side PhA for discharge ADL independence differed between upper and lower limbs in women and men. This finding suggests that regional PhA measurements may provide additional insights into functional status beyond whole-body PhA. The higher accuracy of upper limb PhA in women likely reflects the greater contribution of upper extremity function to ADL tasks like dressing, grooming, and feeding.41) Conversely, lower limb PhA was more predictive in men, potentially due to the importance of ambulation and transfer abilities influenced by lower extremity strength.42) These sex-specific differences may arise from distinct body composition characteristics, with men exhibiting higher muscle mass concentrated in the lower body.43) Assessing regional PhA could enhance discharge planning by identifying specific functional deficits requiring targeted rehabilitation.
Site-specific PhA provided significant incremental prognostic value when added to clinical models. When PhA variables were added to a multivariable logistic regression model that included commonly used clinical predictors, the overall discriminative ability of the model improved significantly. These results suggest that PhA (particularly whole-body measurements) may provide complementary prognostic information that is not fully captured by conventional variables such as age, stroke type, or functional status at admission. Rather than functioning as a standalone biomarker, PhA appears to refine risk stratification by capturing subtle aspects of cellular health and muscle quality that contribute meaningfully to the likelihood of functional independence at discharge.
PhA-assisted collaboration among rehabilitation, nutritional support, and oral management may further improve outcomes. The assessment of site-specific PhA, especially those of the non-paretic limbs, could offer valuable insights into the potential for functional recovery in post-stroke rehabilitation patients. Moreover, considering the high prevalence of sarcopenia and malnutrition in this population,44,45) PhA measurement could serve as a useful indicator for evaluating these conditions. Integrating phase angle assessment with tailored collaboration of rehabilitation, nutrition support, and oral management may further enhance functional recovery and ADL.46,47) Future research should investigate the potential benefits of PhA-guided, individualized rehabilitation nutrition interventions in improving functional outcomes and quality of life for post-stroke patients with impairments such as hemiplegia, reduced ADL performance, malnutrition, and sarcopenia.48-50) This comprehensive approach could address the multifactorial causes of disability and promote better long-term outcomes in stroke survivors.
This study has some limitations. Firstly, the retrospective design and single-center setting may restrict the generalizability of the findings to other healthcare settings. Secondly, although we adjusted for several potential confounding factors, the influence of unmeasured variables, such as dietary intake, physical activity levels, and rehabilitation intensity, cannot be excluded. Thirdly, the use of BIA for assessing phase angles and body composition may be susceptible to measurement errors, particularly in patients with fluid imbalances or altered hydration status. Future prospective, multicenter studies incorporating comprehensive assessments of nutritional status, physical function, and rehabilitation regimens are warranted to validate and expand upon our findings.
In conclusion, this study identified optimal cutoff values of regional PhA for predicting ADL independence at discharge in post-stroke patients. The diagnostic accuracy of non-paretic limb PhA differed between sexes. Moreover, site-specific PhA provided significant incremental prognostic value when added to clinical models. These findings highlight the potential utility of segmental PhA assessment in tailoring the triad of rehabilitation, nutrition support and oral management strategies to enhance functional recovery after stroke.
Notes
CONFLICT OF INTEREST
The researchers claim no conflicts of interest.
FUNDING
None.
AUTHOR CONTRIBUTIONS
Conceptualization, YY, HW, FN, AM, SS, AS, YK, TB, TH, KY; Data curation, YY, HW, FN, AM, SS, AS, YK, TB, TH, KY; Investigation, YY, FN, AM, SS, AS, YK, TB, TH, KY; Methodology, YY, FN, AM, SS, AS, YK, TB, TH, KY; Formal analysis, YY, FN, AM, SS, AS, YK, TB, TH, KY; Writing_original draft, YY, HW, FN, AM, SS, AS, YK, TB, TH, KY; Writing_review & editing, YY, HW, FN, AM, SS, AS, YK, TB, TH, KY.
SUPPLEMENTARY MATERIALS
Supplementary materials can be found via https://doi.org/10.4235/agmr.25.0028.
Incremental discrimination of a clinical-baseline model after adding each continuous site-specific phase angle
Spearman correlation matrix among five site-specific phase angle measurements