Cardiac Manifestations of Sarcopenia

Abstract

Objectives

Screening questions for sarcopenia used in the community (SARC-F) may be regarded as indicators of exercise tolerance.

Design

Observational study.

Setting

We tested the hypothesis that community-living older people who are screened positive for sarcopenia using the SARC-F tool but without a history of heart failure (HF) have a higher prevalence of cardiac abnormalities compared with those who are SARC-F negative.

Participants

Participants were recruited from a territory-wide primary care needs assessment for older people based in community centres, and from non-acute hospitals in the same region as the study centre.

Measurements

Participants with a total score of >=4 and who did not have any history of HF were invited to attend for further cardiac assessment. Grip strength, walking speed, and the 6-minute walk test and echocardiography were carried out. Patients with frailty and at least Grade II diastolic dysfunction were considered to have heart failure with preserved ejection fraction (HFpEF) if they also had concomitant elevated N-terminal prohormone of B-type natriuretic peptide (NT-proBNP) of at least 300 pg/ml.

Results

Diastolic dysfunction (DD) was significantly associated with SARC-F score >=4 and higher circulating NT-proBNP levels. ROC curves evaluating the predictive values of SARC-F, HGS and gait speed for DD showed that a combination of SARC-F and HGS or gait speed provided significant incremental value in predicting DD.

Conclusions

Community living older people with sarcopenia detected using a simple questionnaire have a higher prevalence of DD accompanied by elevated NT proBNP. Addition of hand grip strength or walking speed improve the magnitude of the association. SARC-F may be used as a tool to detect early cardiac dysfunction in the community.

Introduction

Since the introduction of the term ‘sarcopenia' by Rosenberg to describe the phenomenon of age related loss of muscle mass (1), sarcopenia has been included in the ICD-10 code as a disease entity (M62.84) in 2016, with consensus definitions that include muscle strength and physical performance measures. Due to ethnic variations in body size, there are slight differences in definitions between Caucasians and Asians, particularly with respect to cut-off values based on lowest population centiles (2, 3, 4, 5). Rapid screening tools have also been developed and validated, particularly for community-living older people, as a first step in a step care approach to case finding followed by interventions (6).

Due to accelerating muscle loss particularly after age 70 and documented adverse outcomes such as increased mortality, mobility limitations, falls and fractures, poor quality of life as well as metabolic consequences, sarcopenia is of public health importance. Therefore there are calls to include rapid screening tools in the community as part of routine elderly health screening (7), with a view to interventions.

Sarcopenia may also occur as a result of disuse. For example, after prolonged bed rest from short term illnesses, or accompanying chronic diseases limiting mobility such as stroke, or in addition where there is upregulation of inflammatory cytokines such as in chronic obstructive pulmonary disease or heart failure (HF) (secondary sarcopenia). While the development of chronic HF (CHF) increases with age, with approximately equal proportion of HF with reduced ejection fraction (HFrEF) and with preserved ejection fraction (HFpEF), both having poor prognosis (8), the condition is largely undetected in community living older people even though there may be symptoms; perhaps because the symptoms are largely non-specific or attributed to ageing itself. Although there is a literature base on the occurrence of sarcopenia among patients with CHF, the percentage of such publications is small compared with the large volume on CHF itself (9, 10, 11, 12, 13). Sarcopenia is usually discussed as secondary to HF, as a result of reduced exercise capacity, i.e. secondary sarcopenia. However it is possible that in primary sarcopenia, cardiac dysfunction may already be present, due to common underlying pathophysiology underlying both conditions, namely upregulation of inflammatory cytokines, or inflammageing (14). This process is systemic and affects many organs and systems, and underlies many geriatric syndromes of which sarcopenia is one. A consequence of this process on cardiac muscle is increasing stiffness. In addition, metabolic changes associated with obesity, a risk factor for cardiovascular diseases, may also affect muscle function through fat infiltration of skeletal muscle (sarcopenic obesity), while adipocytes increase cytokine release. Thus, a common process may directly lead to both sarcopenia and cardiac remodelling (15).

It is possible that older people living in the community with sarcopenia already have abnormalities in cardiac function even though they have not been diagnosed with CHF. This would not be surprising since some of the screening questions for sarcopenia may be regarded as indicators of exercise tolerance (walking ability; ability to climb stairs). Using data from a health screening project for older people aged 60 years and over, we tested the hypothesis that community-living older people who are screened positive for sarcopenia using the SARC-F tool but without a history of heart failure have a higher prevalence of cardiac abnormalities compared with those who are SARC-F negative.

Method

Participants

Participants were recruited from a territory-wide primary care needs assessment for older people based in community centres using automated data capture followed by computer generated report to guide further action. The project has been described elsewhere (16). Some participants were also recruited from non-acute hospitals in the same region as the study centre. Information from the questionnaire was extracted to construct the SARC-F score (Table 1): Strength- how much difficulty do you have in lifting and carrying 10 lbs (None=0; Some=1; A lot or unable=2); assistance in walking- how much difficulty do you have walking across a room (none=0; some=1; a lot, use aids, or unable=2); rise from a chair- how much difficulty do you have transferring from a chair or bed (none=0;some=1; a lot or unable without help=2); climb stairs- how much difficulty do you have climbing a flight of 10 stairs (none=0; some=1; a lot or unable=2); falls- how many times have you fallen in the past year (none=0; 1–3 falls=1; >=4 falls=2). Participants with a total score of >=4 were classified as having sarcopenia (6). All participants who did not have any history of HF were invited to attend a regional acute hospital (the study centre) as well as a non-acute hospital in the same region for further cardiac assessment. The sample was stratified according to frailty classification, with approximately 100 per frailty group, as described in a previous study (17). Medical follow-up was offered to all participants if abnormalities in cardiac function were detected, either at the same hospital or referred to hospitals nearer their residential address. Written informed consent was obtained from all participants. The study was approved by the Chinese University of Hong Kong-New Territories East Cluster Clinical Research Ethics Committee.

Table 1.

SARC-F screen for sarcopenia

Component	Question	Scoring
Strength	How much difficulty do you have in lifting and carrying 10 pounds?	None = 0 Some = 1 A lot or unable = 2
Assistance in walking	How much difficulty do you have walking across a room?	None = 0 Some = 1 A lot, use aids, or unable = 2
Rise from a chair	How much difficulty do you have transferring from a chair or bed?	None = 0 Some = 1 A lot or unable without help = 2
Climb stairs	How much difficulty do you have climbing a flight of 10 stairs?	None = 0 Some = 1 A lot or unable = 2
Falls	How many times have you fallen in the past year?	None = 0 1–3 falls = 1 4 or more falls = 2

Assessments

Questionnaire

Information obtained by questionnaire include sociodemographic information, smoking and drinking habit, history of cardiovascular disease and diabetes, and any symptoms of HF according to the New York Heart Association classification. Cardiovascular medications used were recorded.

Physical performance measures

Grip strength was measured using a dynamometer (Jamar Hand Dynamometer 5030JO; Sammons Preston Inc., Bolingbrook, IL, USA). Two readings were taken from each side, and the maximum value of all the measurements was used for analysis. Walking speed was measured using the best time in seconds to complete a walk along a straight line 6 metres long. A warm up period of less than 5 minutes was followed by two walks, and the best time was recorded. The 6-minute walk test was conducted in a 20-metre corridor on a flat hard surface, with every 5 metres marked on the side of the walkway. A cone was placed at each end of the walkway to indicate where the participant had to turn. Participants were encouraged to walk continuously for six minutes, and the distance covered was recorded.

Cardiac function assessment: echocardiography

Study participants underwent transthoracic echocardiographic (TTE) examination using a General Electric VIVID E9 ultrasound system. Left ventricular ejection fraction (Simpson's biplane method) and standard diastolic parameters were measured as per American Society of Echocardiography recommendations (18). Briefly, mitral inflow E and A velocities, tissue Doppler e', deceleration time, left atrial volume index, and tricuspid regurgitation, if available, were considered for grading (0, I–IV) of diastolic dysfunction.

Cardiac function assessment: NT-proBNP level measurements

Patients with frailty and at least Grade II diastolic dysfunction were considered to have HFpEF if they also had concomitant elevated N-terminal prohormone of B-type natriuretic peptide (NT-proBNP) of at least 300 pg/ml (19). Blood specimens were collected in Vacuette Z Serum Separator Clot Activator blood collection tubes (Greiner Bio-One International GmbH, Austria). NT-proBNP levels were measured using the Roche Elecsys e411 analyser in a clinical laboratory at Prince of Wales Hospital, Hong Kong that used Roche-supplied standard reagents and followed a protocol for in vitro diagnosis.

Statistical analysis

Normally distributed continuous variables were reported as means + standard deviations (SD), and one-way ANOVA with least significant difference post-hoc test was performed for between-group comparisons. Continuous variables with non-parametric distribution were expressed as median (interquartile range) and compared using Kruskal-Wallis H Test. Categorical variables were expressed as frequencies, and Chi-square or Fisher’s exact test was used for comparison between groups where appropriate. Logistic regression was used to test for associations between binary variables, and the effect sizes were presented as odds ratios (OR) between groups when comparing individual components of SARC-F scale and diastolic dysfunction. Diastolic dysfunction was categorised as a binary variable based on American Society of Echocardiography recommendations (18). The cut-point of 75 years was used in the sub-analysis of age. A two-sided P <0.05 was considered statistically significant. IBM SPSS Version 24.0 (Armonk, New York, LSA) software was used for statistical analyses. DeLong’s test was used to examine statistical difference between ALCs for different ROC curves.

Results

Of 306 older people who participated, 102 had a SARC-F score >=4 (Table 2). These were older, more likely to be men, with higher NYHA class, of lower socioeconomic class as indicated by housing type, less likely to be current alcohol drinkers, and with higher prevalence of self- reported co-morbidities. They also had lower ratio of grip strength adjusted for body mass index (GS/BMI) (0.58±0.16 vs. 0.85±0.29, P<0.01), gait speed (0.66±0.20 vs. 0.96±0.19 m/sec, P<0.01) and 6-minute walk distance (6MWD) (236.80±82.69 vs. 371.80±69.91 m, P<0.01). Atrial fibrillation was more often observed in this group (6.9% (7/102) vs. 2.0% (4/204), P<0.05).

Table 2.

Baseline characteristics of participants by sarcopenia states

Variables	Overall (n=306)	Positive SARC-F (n=102)	Negative SARC-F (n=204)	P value
Socioeconomic demographics
Age	74.73±7.76	79.69±7.56	72.25±6.59	<0.01
Male gender (%, n)	31.0% (95/306)	14.7% (15/102)	39.2% (80/204)	<0.01
BMI	25.26±3.71	25.29±4.09	25.25±3.51	N.S.
NY HA class (%, n)				<0.01
Class I	28.8% (88/306)	2.9% (3/102)	41.7% (85/204)
Class II	37.9% (116/306)	27.5% (28/102)	43.1% (88/204)
Class III	31.0% (95/306)	62.7% (64/102)	15.2% (31/204)
Class IV	2.3% (7/306)	6.9% (7/102)	0
Housing types (%, n)				<0.01
Public estate	44.8% (137/306)	57.8% (59/102)	38.2% (78/204)
House ownership	23.2% (71/306)	15.7% (16/102)	27.0% (55/204)
Private estate	27.1% (83/306)	20.6% (21/102)	30.4% (62/204)
Others	4.9% (15/306)	5.9% (6/102)	4.4% (9/204)
Living alone	25.2% (77/306)	38.2 (39/102)	18.6% (38/204)	<0.01
Smoking status				N.S.
Current smokers	2.6% (8/306)	2.9% (3/102)	2.5% (5/204)
Previous smokers	19.0% (58/306)	11.8% (12/102)	22.5% (46/204)
Non-smokers	78.4% (240/306)	85.3% (87/102)	75.0% (153/204)
Alcohol status				<0.01
Current drinkers	20.3% (62/306)	5.9% (6/102)	27.5% (56/204)
Previous drinkers	14.7% (45/306)	13.7% (14/102)	15.2% (31/204)
Non-drinkers	65.0% (199/306)	80.4% (82/102)	57.4% (117/204)
Comorbidities
Hypertension	65.7% (201/306)	76.5% (78/102)	60.3% (123/204)	<0.01
Diabetes mellitus	28.1% (86/306)	35.3% (36/102)	24.5% (50/204)	<0.05
Ischaemic heart disease	6.5% (20/306)	10.8% (11/102)	4.4% (9/204)	<0.05
Atrial fibrillation	2.9% (9/306)	4.9% (5/102)	2.0% (4/204)	N.S.
Previous myocardial infarction	3.0% (9/306)	4.9% (5/102)	2.0% (4/204)	N.S.
Valvular heart disease	7.2% (22/306)	6.9% (7/102)	7.4% (15/204)	N.S.
Chronic pulmonary disease	11.1% (34/306)	19.6% (20/102)	6.9% (14/204)	<0.01
Chronic arthropathy	42.2% (129/306)	62.7% (64/102)	31.9% (65/204)	<0.01

NYHA, New York Heart Association; ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; CCB, calcium channel blocker; N.S., Not significant

With respect to echocardiographic findings, there was no difference in left ventricular ejection fraction (LVEF) between the two groups. Those with SARC-F >=4 had more diastolic dysfunction (DD) (Table 3). DD and Grade II–IV DD with preserved LVEF were significantly more prevalent in positive SARC-F group (DD: 81.3% (78/96) vs. 57.9% (114/197), P<0.01; Grade II–IV DD: 29.2% (28/96) vs. 12.7% (25/197), P<0.01). These participants also had significantly higher circulating NT-proBNP levels (Log transformed): 2.32±0.54±0.48 vs. 1.96±0.42, P<0.01) pg/ml.

Table 3.

Echocardiographic parameters between sarcopenia states

Variables	Overall (n=306)	Positive SARC-F (n=102)	Negative SARC-F (n=204)	P value
Systolic variables
LVEF	61.39±7.95	6074±8.34	6171±774	N.S.
LV septal s' (m/s)	0.058±0.013	0.056±0.013	0.060±0.014	<0.05
LV lateral s' (m/s)	0.069±0.018	0.065±0.017	0.071±0.019	<0.01
RV lateral s' (m/s)	0.114±0.025	0.115±0.028	0.113±0.024	N.S.
Diastolic variables
LAVI (ml/m2)	27.77±10.76	31.40±10.08	25.92±10.65	<0.01
E wave (m/s)	072±0.19	077±0.19	0.69±0.18	<0.01
A wave (m/s)	0.82±0.22	0.90±0.23	079±0.20	<0.01
E/A	0.90±0.28	0.86±0.27	0.91±0.29	N.S.
LV septal e' (m/s)	0.053±0.15	0.048±0.014	0.055±0.014	<0.01
LV lateral e' (m/s)	0.068±0.21	0.062±0.020	0.071±0.021	<0.01
Average e' (m/s)	0.060±0.017	0.054±0.017	0.062±0.018	<0.01
LV septal Ele'	14.59±5.44	17.27±5.90	13.24±4.66	<0.01
LV lateral Ele'	11.45±5.88	13.61±5.52	10.36±4.00	<0.01
Average Ele'	12.50±4.66	14.59±5.12	H.46±4.04	<0.01
DT (ms)	187.98±40.84	191.27±43.81	186.29±39.24	N.S.

LVEF, left ventricular ejection fraction; LAVI, left atrial volume index; DT, deceleration time; N.S., Not significant

In univariate logistic regression analysis, SARC-F >=4 was associated with increased Odds Ratio of DD by 3.16- (95% CI: 1.76–5.67, P<0.01) and Grade II–IV DD by 2.83-fold (95% CI: 1.54–5.20, P<0.01). The association between sarcopenia states and DD remained after adjusting for comorbidities (DD: adjusted OR: 2.70; 95% CI: 1.47–4.96, P<0.01; Grade II–IV DD: 2.61; 95% CI: 1.38–4.94, P<0.01) but not after adjustment for age and gender using multivariate logistic regression models (DD: adjusted OR: 1.36; 95% CI: 0.69–2.67, P>0.05; Grade II-

IV DD: adjusted OR: 1.57; 95% CI: 0.78–3.16, P>0.05).

SARC-F >=4 was also associated with increased OR of NT-proBNP >125pg/ml by 3.29- (95% CI: 2.00–5.40, P<0.01), and >300pg/ml by 3.43-fold (95% CI: 1.93–6.09, P<0.01) in univariate logistic regression, and remained after adjusting for comorbidities but not for age and gender.

When individual components of the SARC-F were examined, with the exception of the question ‘fall(s) within 1 year', the other 4 components including ‘carrying 10 pounds', ‘walking across the room', ‘climbing 10 steps' and ‘rising from bed or chair' were all associated with increased OR of DD by 2.38–8.80 fold. The associations between these SARC-F components and DD remained after adjusting for comorbidities. Only the OR for the question ‘climbing 10 steps' remained significant after additional adjustment for age and gender.

ROC curves were developed to evaluate the predictive values of SARC-F, HGS and gait speed for DD with preserved LVEF (Figure 1). Combining SARC-F and HGS or gait speed provided a significant incremental value in predicting DD with preserved LVEF (SARC-F+HGS: AUC: 0.882; 95% CI: 0.815–0.949, P<0.01; DeLong test: SARC-F+HGS vs. SARC-F: P<0.01; SARC-F+gait speed: AUC: 0.829; 95% CI: 0.748–0.909, P<0.01; DeLong test: SARC-F+gait speed vs. SARC-F: P<0.05) A combination of SARC-F, HGS, age, gender, and comorbidities did not significantly improve the predictive value for DD with preserved LVEF: SARC-F+HGS+age+gender+comorbidities vs. SARC-F+gait speed or SARC-F + HG: P>0.05. (Curve not shown in Figure 1)

Figure 1.

Receiver operating characteristic (ROC) curve of SARC-F in detecting advanced cardiac dysfunction with preserved LVEF

Model 1: SARC-F alone; Model II: SARC-F + gait speed; Model III: SARC-F+HG

LVEF, left ventricular ejection fraction; LAVI, left atrial volume index; DT, deceleration time; N.S., Not significant

Discussion

The findings support our hypothesis, that among community-living older people, those with a SARC-F score >=4 suggestive of sarcopenia had higher prevalence of diastolic dysfunction (grade>I), higher prevalence of atrial fibrillation, and NT-pro BNP, after adjusting for co-morbidities. Of all the constituent questions of SARC-F, the four that may be regarded as related to exertion are associated with SARC-F, while the remaining question about falls history, an indicator of weakness, was not. Furthermore, a combination of SARC-F score combined with an objective measurement that is included in the definition of sarcopenia such as walking speed, improves the strength of that association, which is only marginally improved by the addition of age, gender co-morbidities, and strength measure. It may be argued that the use of SARC-F as a screening measure for sarcopenia may also be a method of screening for undiagnosed heart failure among community-living older people, since some of the questions may be regarded as indicators of exercise tolerance, which is a symptom of heart failure. However the association appears to be with heart failure of preserved ejection fraction (HFpEF) rather than with systolic dysfunction-HFrEF (heart failure with reduced ejection fraction).

This finding may be expected when possible common underlying pathophysiological mechanism are considered. Chronic inflammation with ageing, or inflammageing, has been proposed to be a causal factor, as a result of central obesity, increase gut permeability, changes to gut microbiota, cellular senescence, NLRP3 inflammasome activation, mitochondrial dysfunction, immune cell dysregulation, and chronic infections (14, 20). Inflammation leads to cardiac hypertrophy and stiffness, contributing to cardiac remodeling and dysfunction with increased risk of HFpEF (15). However, ageing per se may result in progressive deterioration in cardiac structure and function, resulting in left ventricular hypertrophy, impaired diastolic function but relatively preserved systolic function, reduced exercise capacity, and increased prevalence of atrial fibrillation, independent of conventional risk factors for heart disease. Such changes result in the phenotype of HFpEF, which is especially prevalent in aged women (21).

The findings of this study support the recently proposed concept that HFpEF may be regarded as a geriatric syndrome (22, 23), similar to other syndromes such as sarcopenia and frailty. In fact, sarcopenia has been described as physical frailty (24). Similar to studies on the association between sarcopenia and HF, the majority of studies relating to frailty and HF documents adverse outcomes in patients with HF between those who are frail and those who are not, with few studies on community-living subjects with no self reported diagnosis of HF. Separately using the same cohort, we had documented that there was a gradation in DD and NT-proBNP levels among the three frailty classes, with the highest prevalence of DD in the frail group (17). From the public health or primary care point of view, there are already initiatives for rapid community screening of older people to include geriatric syndromes (16, 25), and the findings of this study add to the implications of screening for sarcopenia in that a cardiac assessment may be indicated as a further step as part of a step care approach to primary care of older people. This is not futile as there are interventions for sarcopenia in terms of nutrition and exercise (26), and currently treatment of HFpEF is also exercise rather than pharmacological (27). In future it would be of interest to see if drugs effective for sarcopenia may also be of benefit for HFpEF.

There are limitations in this study. We relied on the SARC-F screening tool to classify participants into those with and without sarcopenia. Diagnoses according to consensus panel definitions usually include objective strength, physical performance, as well as muscle mass measurements. We did not carry out any muscle mass measurements; only grip strength and physical performance measures. These were examined in the receiver operating characteristic curves for association with DD, revealing incremental predictive value. In practice, one may recommend further cardiac assessment using a combination of SARC-F and either strength or physical performance measures. Currently muscle mass assessment requires equipment, and neither dual energy X-ray absorptiometer nor bioimpedance provide an accurate measure of muscle mass. Hence, a meta-analysis of harmonised studies counterintuitively have found that muscle mass alone does not predict any adverse outcomes (Unpublished data presented at the Sarcopenia Definition and Outcomes Consortium meeting, Boston, USA, November 2018). This approach may be relevant for detecting undiagnosed HF in the community, since one does not have to rely on routine echocardiography or blood tests.

Furthermore, this study may not be used as an indicator of the prevalence of sarcopenia in the community, and hence indirectly that of HFpEF, because of the purposeful sampling of approximately one third of participant for each of the three categories of frailty. Since sarcopenia had been considered to represent physical frailty, it is likely that the prevalence in this sample would be higher and not representative of community-living older people as a whole. In fact, this purposeful sampling of large enough numbers of people who have sarcopenia may be considered a strength of the study, allowing balanced analyses to be carried out. To date, the authors are not aware of a similar study designed to address the research question posed. This study paves the way for future research into examining whether interventions for sarcopenia may also affect DD.

Conclusions

In conclusion this study shows that community living older people aged 60 years and over with sarcopenia detected using a simple questionnaire of five questions have a higher prevalence of DD accompanied by elevated NT proBNP, and that addition of hand grip strength or walking speed improve the magnitude of the association with an AUC of 0.828. SARC-F may be used as a tool to detect early cardiac dysfunction in the community.

Conflict of Interest

None declared.

Ethical Standards

Written informed consent was obtained from all participants, and the study was approved by the Chinese University of Hong Kong Clinical Research Ethics Committee.