Prevalence and factors associated with sarcopenia among older adults in a post-acute hospital in Singapore

Background: Sarcopenia is common in older adults worldwide, but its prevalence varies widely owing to differences in diagnostic criteria, population sampled, and care setting. We aimed to determine the prevalence and factors associated with sarcopenia in patients aged 65 and above admitted to a post-acute hospital in Singapore. Methods: This was a cross-sectional study of 400 patients recruited from a community hospital in Singapore. Data including socio-demographics, physical activity, nutritional status, cognition, clinical and functional status, as well as anthropometric measurements were collected. Sarcopenia was defined using the Asian Working Group for Sarcopenia 2019 criteria (AWGS2019). Results: Of the 383 patients with complete datasets, overall prevalence of sarcopenia was 54% while prevalence of severe sarcopenia was 38.9%. Participants with increased age, male gender and a low physical activity level were more likely to be sarcopenic, while those with higher hip circumference and higher BMI of [≥]27.5m/kg2 were less likely to be sarcopenic. Other than the above-mentioned variables, cognitive impairment was also associated with severe sarcopenia. Conclusions: There is a high prevalence of sarcopenia in older patients admitted to a post-acute hospital in Singapore. There is an urgent need to address this important clinical syndrome burden and to identify patients at higher risk of sarcopenia in post-acute settings in Singapore for early intervention.

There is a high prevalence of sarcopenia in older patients admitted to a post-acute hospital in Singapore.There is an urgent need to address this important clinical syndrome burden and to identify patients at higher risk of sarcopenia in post-acute settings in Singapore for early intervention.Word count "abstract": 213 .
. For example, sarcopenia prevalence in a community sample of individuals aged 21-90 years old compared to those aged 60 years and above was 13.6% and 32.2% respectively.(21) Another study reported sarcopenia prevalence of 76% among community ambulant adults aged 65 and above who were at medium or high risk of malnutrition.(22) The prevalence of sarcopenia in primary care and specialist outpatient clinics ranged from 27.4% in patients aged 60-89 years old with Type 2 Diabetes Mellitus using AWGS criteria, to 44% in patients aged 65 and above based on SARC-F questionnaire only.(19,25) However, till date there is none investigating sarcopenia prevalence using established recommended criteria in the inpatient setting.We aim to determine the prevalence of sarcopenia, severe sarcopenia and their associations in an inpatient post-acute hospital setting in Singapore.

Setting
This cross-sectional study was conducted from May to November 2022 at Outram Community Hospital (OCH).A community hospital in Singapore is a purpose built hospital to provide medical, nursing and rehabilitation care for patients who require a short period of continuation of care after their stay in the acute hospitals i.e. post-acute care before being discharged into the community.(26) Admissions into OCH come from the Singapore General Hospital (SGH), the largest colocated acute tertiary hospital in Singapore servicing the southeastern region of Singapore. .

Study participants and recruitment
All patients aged 65 years old and above admitted to OCH under subacute or rehabilitation service from 17 May to 16 November 2022 were consecutively screened based on eligibility criteria.Participants who were unable to understand English or Mandarin, refused to participate, or unable to give informed consent or follow instructions due to conditions such as neuropsychiatric or neurocognitive disease were excluded.Additional excluded participants included those who were terminally ill; medically unstable; had conditions which precluded sarcopenia assessment such as anyone with a cardiac pacemaker, an implantable defibrillator, amputation of limb(s), intravenous hydration, fluid overload and had temporary restrictions in weightbearing status of upper or lower limbs.Eligible participants were then recruited after written informed consent was obtained.All participants recruited were assigned to a unique serial number in order to delink and deidentify them.All data was collected during the inpatient stay in OCH, with no further follow-up or contact after discharge.This study was approved by SingHealth Centralized Institutional Review board (CIRB Ref 2021/2817).

Data Collection
All interviewers received standardized training on interview techniques with trial interviews held with trainer to ensure fidelity.Face-to-face interviews with the patients were conducted in Mandarin and English.
. Data collected included 1) demographics (age, sex, race, marital status, living setup, type of dwelling and highest educational qualification); 2) physical activity and sedentary time; 3) nutritional status; 4) cognition; and 5) other clinical parameters (weight, height, smoking history, Modified Barthel Index (MBI) scores on admission, Charlson Comorbidity Index (CCI); history of COVID-19 infection; waist circumference (WC); hip circumference (HC) and Rehabilitation Diagnostic Groups (RDGs) classification, an administrative framework to group patients admitted for rehabilitation so as to facilitate transition and collaboration across the stakeholders in the care spectrum.Physical activity level and sedentary time were assessed using the Global Physical Activity Questionnaire (GPAQ), which was developed by WHO with standardized approach and question guide available both in English and Mandarin translations, and used in multiple Singapore studies including the National Health Survey 2010 and 2019.(21,(27)(28)(29)(30)(31)(32) As per the GPAQ, participants' sedentary time in minutes per day and their physical activity level (low and moderate or high) were collected.(33) The 3-Minute Nutritional Screening (3-MinNS) tool, developed in Singapore and validated for medical and surgical inpatients in a Singapore hospital, was used to screen nutritional status.(34,35) Participants were grouped into low (0-2), moderate (3)(4), or severe (5-9) malnutrition risk.(34) Cognitive assessment was done using the Chinese Mini Mental State Examination (CMMSE), which was also validated locally in English and Mandarin.(36) Those with CMMSE scores of 23 and below were considered as having impaired cognition.(37) .The participants' inpatient medical records were accessed to retrieve information regarding the following: MBI on admission; CCI; history of COVID-19 infection; RDGs; weight (in kilogrammes) and height (in metres) on admission.Body mass index (BMI) was calculated as body weight divided by square of height (kg/m 2 ).Cut-offs were based on WHO BMI risk categories for cardiovascular disease and diabetes in Asian populations.(38) Anthropometry Waist circumference (WC) was measured using a stretch-resistant tape at the midpoint between the lower margin of the least palpable rib and the top of the iliac crest.Waist circumference cut-offs were determined as per The International Diabetes Federation (IDF) consensus worldwide definition for metabolic syndrome.(39) Hip circumference (HC) was measured around the widest portion of the buttocks.(40) Waist-hip ratio (WHR) was calculated by dividing WC over HC.Cut-offs for WHR were based on World Health Organization (WHO), redefining obesity -the Asia-Pacific perspective.(41) Sarcopenia assessment There have been several proposed diagnostic algorithms for sarcopenia.In 2010, the European Working Group on Sarcopenia in Older People (EWGSOP) proposed the first practical clinical definition and diagnostic criteria for sarcopenia based on assessment . of muscle mass, muscle strength and physical performance.(42) This was revised in 2018 to EWSOP2, which used low muscle strength as the primary parameter of sarcopenia.(10) To address differences in cut-off values of measurements in Asian populations from Europeans, an Asian consensus was derived by the Asian Working Group for Sarcopenia (AWGS) in 2014 and revised in 2019.(43,44) In our study, sarcopenia was assessed and diagnosed using criteria as per AWGS2019 (44).Sarcopenia was diagnosed in the presence of low muscle mass, with either low muscle strength or low physical performance (PP).Those with low muscle mass, low muscle strength, and low PP, were further subclassified as "severe sarcopenia" for sub-group analyses.(44) Muscle mass Appendicular skeletal muscle mass (ASM) was determined using a multifrequency bioelectrical impedance analysis (BIA) InBody S10 Body Composition Monitor.ASMI was calculated by dividing ASM by height squared (kg/m 2 ).BIA measurements were conducted under standardised protocols i.e. before therapy sessions, in supine position with limbs abducted and ensuring no contact with metal frame of bed.(45) Low muscle mass was defined as appendicular skeletal muscle mass index (ASMI) <7kg/m2 in males and <5.7kg/m2 in females.

Muscle strength
Muscle strength was assessed via handgrip strength (HGS) which was measured using a dynamometer (BASELINE 12-0240 standard hydraulic hand dynamometer) in seated posture, with shoulder adducted and elbow flexed to 90 degrees and forearm in neutral as recommended by American Society of Hand Therapists.(46) The maximum .reading from at least two trials using either hand in a maximum-effort isometric contraction was used for analysis.(44) Low muscle strength was defined as handgrip strength (HGS) <28kg in males and <18kg in females.

Physical Performance
Low PP was defined as gait speed <1.0m/s over 6-metre walk (6mGS), or 5-time chair stand test (5CST) requiring ≥12s, or Short Physical Performance Battery (SPPB) score of ≤9.6mGS was measured as time taken to walk 6m at a normal pace from a moving start, without deceleration, with average of two trials taken as the recorded speed (m/s) (44).Walking aids were permitted if necessary and documented if it was used.5CST was measured as the time taken to rise and sit five times as quickly as possible with no contact against the back of the chair and maximal extension of knees(47).SPPB is a 3-part performance-based test of balance, gait speed, and 5CST, with a score of 0 to 12.

Statistical analyses
Categorical variables were presented as proportions and continuous variables summarized as means+/-SD or medians with interquartile ranges [IQR (25th percentile, 75th percentile)] as appropriate.Pearson chi-square test was used to compare categorical variables and logistic regression to compare continuous variables as appropriate for univariate analyses.outcomes.Similarly multinomial regression was done on all variables to calculate adjusted odds ratio associated with severe sarcopenia outcomes.All analyses were two-tailed and p value <0.05 was considered statistically significant.Statistical analysis was performed using Stata Version 17 (StataCorp, College Station, TX, USA).

Results
A total of 400 patients were recruited, of which 17 (4%) had incomplete data due to discharge or transfer to acute hospital for acute medical issues.Hence, 383 (96%) patients with complete data were analysed in this study.Figure 1 shows the process of identifying sarcopenia in our study population using AWGS2019 diagnostic algorithm.Prevalence of sarcopenia and severe sarcopenia Prevalence of sarcopenia using AWGS 2019 criteria in our study population was 54.0% (n=207) with 66.4% of males being sarcopenic and 45.9% of females being sarcopenic.Prevalence of severe sarcopenia in our cohort was 38.9% (n=149).

Characteristics of study population and unadjusted odds ratio
Table 1 illustrates the characteristics of our participants with corresponding unadjusted odds ratios of sarcopenia.The median age of participants was 75 years old (IQR: 70-81), with more females (60.3%) than male participants, and majority (93%) were Chinese.Slightly more than half (56.1%) were currently married.Majority (91.9%) live in public governmental housing (Housing Development Board (HDB) apartments).About two thirds (60.4%) of our cohort have less than secondary school education and .only 20% of participants were smokers or ex-smokers.2/3 of our participants had CCI more than "0".62% of participants were not known to be infected by COVID-19 prior to recruitment into study.Slightly more than half (56.4%) were at moderate or high risk of malnutrition.Physical activity level was evenly distributed with half (50.9%) of the participants assessed to have low physical activity level.Approximately half of our participants (49.9%) had CMMSE 23 or less.Slightly over half (57.4%) of participants had BMI of 23kg/m 2 or more, while approximately 2/3 had an elevated waist circumference and waist-hip ratio.Half (54.1%) of the participants had an MBI score of 0-60 on admission, indicating severe to total dependency in activities of daily living.Participants were mostly admitted for rehabilitation (RDG) due to musculoskeletal conditions (48.8%), followed by deconditioning (18.5%), others (14.4%), hip fracture (11.5%), and stroke (5.1%).3).

Discussion
We found a high prevalence of sarcopenia in an inpatient post-acute rehabilitation setting in Singapore.In addition, what is worrisome is that severe sarcopenia is highly prevalent with 1 in 3 inpatients having severe sarcopenia.Our sarcopenia prevalence is in keeping with worldwide literature where such post-acute units have the highest prevalence of sarcopenia followed by nursing home residents, then hospitalized older adults and lowest in community dwelling older adults.(48) This confirms the burden of disease in units performing rehabilitation and suggests that patients should be intervened earlier upstream even before they are being referred or admitted to such settings.There is also a certain urgency to identify this group of patients for further study and conduct interventions with the aim to reverse the sarcopenia.Like many studies, we found that ageing is associated with sarcopenia.(21)(22)(23)49) This is in keeping with the observation where older adults lose muscle mass with increasing age with intramuscular and intermuscular fat infiltration.(50-53)Many underlying hypotheses for this had been proposed including role of chronic inflammation, intracellular biochemical pathways, concurrent metabolic chronic diseases more prevalent in older adults leading to the onset of sarcopenia.(54,55) However, interesting developments looking at birth weights and later onset sarcopenia seem to hint at other factors that go beyond ageing as contributory factors for sarcopenia, setting the stage for a comprehensive review at reversing sarcopenia at all levels.(56,57) .Most global literature found a significant association between males and sarcopenia like our study although there are some studies showed an association between females and sarcopenia instead.(18,19,21,22) Nonetheless, this highlights likely unique gender differences in developments of sarcopenia.(21,58) It has been postulated that myostatin causing catabolism in males and reduced IGF-1 leading to anabolic decline in females have a role in the sex differences in sarcopenia development.(59,60) Other clinical parameters such as men losing muscle strength and testosterone faster than women also support the association of men being at higher odds of sarcopenia than women.(61-65)Future trials are needed to formulate sex specific interventions tailored to reverse sarcopenia so as to conclusively determine the causal effects of these associations.Consistent with literature, low physical activity has an inverse association with muscle mass and higher odds of sarcopenia.(66-68)What remains debatable is designing an evidence-based exercise regime to reverse sarcopenia.The differential role of aerobic, resistance training and a combination of the types of physical activity may play different roles in preventing or reversing sarcopenia.(69) Incorporating the right mix for the best effect is currently a knowledge gap worth studying in the future.Higher hip circumference being a surrogate measure of gluteal musculature is understandably inversely associated with sarcopenia.(19,(70)(71)(72) In addition, our study finds that a higher BMI is also protective of sarcopenia similar to other studies.(18,(73)(74)(75) There may be a few explanations for the above.Firstly, a higher BMI may translate to increased load-bearing requirements during daily activities increasing muscle mass.(76)Secondly, a higher BMI may be due to better nutritional intake with a study .showing associations between BMI and total protein intake.(77)Thirdly, BMI is likely an inadequate surrogate measure of adiposity in older adults.(78,79) In fact, sarcopenic obesity (SO) prevalence defined with BMI underestimates those defined by body fat percentages in the same cohort in a study.(74) Interestingly, a higher BMI range, confers a paradoxical survival benefit in older adults.(80,81) A study comparing all-cause mortality within the same cohort between groups with BMI v.s.body fat percentage measurements showed no increased mortality risk in apparently obese category of BMI while those defined obese with body fat percentage measurements did.(82) Perhaps it is time to reconsider the utility of previously accepted BMI range predefined in Asians older adults.Despite the limitations of BMI as a surrogate measure of body fat composition, it is undeniably an easily obtainable measure in a busy setting with limited resources.However, there is a research gap in defining what is a normal BMI range correlating to sarcopenic outcomes in long term prospective prognosis studies.It may also be timely to standardise the definition of sarcopenic obesity using body fat percentages norms in future Delphi consensus studies among content experts.However more studies need to be done to correctly classify obesity using body fat percentage adjusted for sex, age and race.Aligning sarcopenic obesity diagnostic criteria will stimulate further research in this area and allow for confident synthesis of data outcomes across different studies.This will pave the way for interventional trials.In our study, cognitively impaired participants have a higher odd of severe sarcopenia although not to sarcopenia while global literature found an association between .cognitive impairment and sarcopenia.(83) We believe the loss of association of cognitive impairment with sarcopenia could be due to type 2 error.Our study did not find an association between categories of nutritional risks and sarcopenia on multivariate analysis although there was a statistically significant association on univariate analysis.Again, this could be due to type 2 error.An alternative explanation is because of the inherent differential property of a screening tool and a full diagnostic assessment of nutritional state.For example, a study showed GLIM assessment of nutrition predicted sarcopenia among malnourished participants while a screening tool failed to do so.(84) To our knowledge, this is the first inpatient study in Singapore to report sarcopenia prevalence using AWGS diagnostic criteria.Our study affirms some of the common associations associated with sarcopenia in line with global literature.This foundational evidence will provide the possibility of developing technology or artificial intelligence driven case finding for sarcopenia in an inpatient setting via a risk calculator.This is especially important as clinicians face challenges with equipment and manpower investiture with current diagnostic criteria of sarcopenia.Our study will also provide the basis for future intervention targeted at risk factors in our population to modify sarcopenic outcomes.Some other strengths of our study include a reasonably large number of recruitments with low dropouts and missing data, as well as the consecutive sampling methodology for all patients admitted to OCH to achieve a more representative sample of the population.

Limitations
. This study has several limitations.Our single site study participants may not be reflective of inpatients in other post-acute settings in other countries.Our findings cannot be generalized to those with participants characteristics that were excluded as per study protocol.Additionally, causal relationship of the associated factors cannot be established as it is a cross-sectional study.Our study estimated appendicular lean mass using the InBody S10 BIA which was not validated locally.However, the use of a multifrequency BIA is an accepted method of measurement by AWGS2019 with gender-specific cut-offs.The team utilised the InBody S10 for its ability to perform measurements in a supine position, enabling assessment among participants with functional limitations.Additionally, utilisation of a BIA machine will allow for comparative data in future interventional studies spanning across inpatient to the community, where access to DXA may be difficult.The Jamar hydraulic hand dynamometer is the recommended brand for assessing handgrip strength, however this study utilised the Baseline hydraulic hand dynamometer instead for logistical reasons as multiple sets were readily available and routinely used by the rehabilitation department therapists, who were the assessors of handgrip for this study.The Baseline dynamometer has been validated against the Jamar dynamometer to measure equivalently and hence can be used interchangeably (85).

Conclusions
Our study found a high level of sarcopenia among inpatient older adults.We also found that participants with increased age, male gender, and a lower physical activity level were more likely to be sarcopenic, while those who had a higher hip circumference and higher BMI of 27.5 and above were less likely to be sarcopenic.In .addition, cognitively impaired participants were more likely to be severely sarcopenic.There is an urgent need to address the burgeoning burden with sarcopenia and identify patients at higher risk of sarcopenia in post-acute settings in Singapore for early intervention.Acknowledgements: Nil Conflict of interest: The authors declare no conflict of interest.Author's contributions: Charmaine Tan was the study's principal investigator.J Aw and Charmaine Tan were responsible for the conception of the study.J Aw, SSYS, Doris L., Ganeshan K and LL Low were the co-investigators.J Aw and Charmaine Tan were responsible for the study design.Charmaine Tan and SSYS were responsible mainly in the coordination of the training of the interviewers.J Aw, Charmaine Tan, XL Low, HY Quek and Andrea Ong took consent from the participants.XL Low, HY Quek and Andrea Ong were responsible for double checking data collected.Doris L., SV Koh and Ganeshan K. were responsible for the measurement of participants for sarcopenia.J Aw, Charmaine Tan, SSYS and Olivia Xia cleaned the data and double checked the data output to ensure validity of the data.J Aw and Charmaine Tan provided the initial draft of the manuscript.J Aw, Charmaine Tan, LL Low, SSYS and Olivia Xia played a role in the interpretation of the data and revision of manuscript critically for important intellectual content and agreed to the final submission.
. Financial disclosure: This study was funded by the 2021 SingHealth Regional Health System (RHS), Population-based, Unified, Learning System for Enhanced and Sustainable Health (PULSES) Grant ID: CGOCt21S06.Funds were used to recruit interns for coordinating administrative duties of the study on a day to day basis with oversight from the P.I. and Co-Is.Part of the fund was used to purchase the BIA InBody S10 and also a laptop that is encrypted to perform data analysis together with acquisition of Stata software licence.

Table 1 :
Characteristics of study population and factors associated with sarcopenia on univariate analysis a Cut-offs as per WHO BMI risk categories for cardiovascular disease and diabetes in Asian populations b Cut-offs as per IDF c Cut-offs as per WHO # p-value required for significance after Bonferroni correction is 0.0083 *missing data n=2 Factors associated with sarcopenia Older age [OR 1.06 (1.01-1.12)],malesex[OR 2.80 (1.12-7.02)]andlowphysical activity [OR 2.13 (1.17-3.89)]wereassociatedwith sarcopenia.Higher BMI more than or equal to 27.5 [OR 0.16 (0.05-0.52)] and greater hip circumference [OR 0.86 (0.81-0.92)] are inversely associated with sarcopenia (Table2).

Table 2 :
Multivariable analysis on variables associated with sarcopenia

Table 3
Multinomial regression on association factors with sarcopenia and severe sarcopenia a cut-offs for waist circumference as per HPB Asia Pacific Consensus and International Diabetes Federation b cut-offs for WHC as per WHO classification of abdominal fatness .c as assessed using WHO GPAQ d Rehab Diagnostic Groups *missing data n=2