Associations of fat mass and muscle function but not lean mass with cognitive impairment: The Yishun Study.

BACKGROUND: Sarcopenia and obesity are reportedly associated with risk of cognitive decline, and sarcopenic obesity (SO) heightens the risk, but the evidence is sparse and inconclusive. This study aimed to examine the association between SO and cognitive impairment.
METHODS: A total of 542 community-dwelling adults aged between 21 and 90 years were recruited. All participants were assessed for body composition (dual X-ray absorptiometry), handgrip strength (HGS), gait speed (GS), and cognitive function (Repeatable Battery for the Assessment of Neuropsychological Status). Sarcopenia was defined by the presence of low appendicular lean mass index (ALMI) and low HGS or low GS according to the 2019 Asian Working Group for Sarcopenia criteria, and obesity was defined based on the upper two quintiles of fat mass index (FMI).
RESULTS: Sarcopenia alone or in combination with obesity were not significantly associated with cognitive impairment after controlling for confounding variables. Obesity on its own was significantly associated with greater odds of impaired attention (OR: 2.05, 95%CI 1.12-3.82). Low ALMI was not associated, but low HGS, slow GS, and high FMI were individually associated with cognitive impairment: low HGS and immediate memory (OR: 1.91, 95% CI 1.04-3.49); low GS and immediate memory (OR: 2.17, 95% CI 1.26-3.72); high FMI and attention (OR: 2.06, 95% CI 1.22-3.51). Co-occurring high FMI with either low HGS or slow GS exacerbated the observed odds of global and domain-specific (attention, visuospatial) cognitive impairment.
CONCLUSIONS: Lean mass is not relevant, whereas muscle strength and physical performance or adiposity are relevant in defining sarcopenia or sarcopenic obesity in terms of their cognitive impacts.

Introduction

Around 50 million people live with dementia worldwide, and this number is projected to increase to 152 million by 2050 [1]. While there is currently no cure for dementia, there is good evidence in support of the prevention or delaying dementia in at-risk individuals in the pre-dementia stage [2]. Mild cognitive impairment is a pre-dementia stage that can be identified years before onset of dementia.

Sarcopenia is a geriatric syndrome, characterised by the progressive loss of skeletal muscle mass and muscle strength with age [3]. With its increasing prevalence in Asia [4], sarcopenia is a public health challenge given its association with functional decline and physical disability [5]. It is also associated with a >2-fold increased risk of cognitive impairment [6,7], plausibly attributed to shared pathophysiology between the two conditions [8,9].

Concurrent with reduction in muscle mass, increase in total body fat and abdominal adiposity is another age-associated phenomenon [10]. Studies have reported mixed findings of a positive relationship between obesity and cognitive impairment [11,12], or a negative relationship [13] depending on age and sex of subjects [14-17]. While obesity in mid-life (45-65y) is an established risk factor for future dementia [2], being obese (body mass index (BMI) ≥30) in late-life (>65y) reduced the risk of dementia compared to normal BMI [18]. This “obesity paradox” in dementia risk across time may be explained by the age-related body composition changes in the trajectory toward sarcopenia and disability [18]. The conflicting evidence and paradox are likely due to the use of BMI to measure obesity because BMI does not differentiate between fat and fat-free mass [19]. Furthermore, most studies only assessed global cognition functioning and did not distinguish between different domains of cognition [14,15].

The coexistence of sarcopenia and obesity—sarcopenic obesity (SO) [20], has been purported to have additive exacerbating effects on cognitive performance [9,21]. However, the evidence on this relationship remained inconclusive. Two studies on community-dwelling older adults found that SO was associated with poorer cognitive performance than either sarcopenia or obesity alone [21,22], while another study reported similar results only among older adults aged 70 and above but not among those aged 60–69 years [9].

The lack of an age appropriate and standardised obesity definition poses a challenge in studying the negative health consequences associated with SO [23]. The aforementioned studies have used percentage body fat [21,22], and waist circumference [9] to determine obesity. Recently, we presented data to show that fat mass index (FMI) may be the most optimal criterion for SO among the obesity measures, as it accounts for body size differences and improves the association between adiposity and physical function [24]. To date, no studies have examined the relationship between cognition and SO based on FMI.

Given the global pandemic of obesity and that sarcopenia could also occur in middle-aged adults [25,26], it is paramount to understand the association between SO and cognitive impairment in a general adult population inclusive of all age groups. Hence, this study aimed to examine the association between SO and global as well as domain-specific aspects of cognition among healthy community-dwelling adults. To better understand this association, this study also examined the association between individual components of SO and cognitive impairment.

Materials and methods

Participants

This is part of the Yishun Study, a cross-sectional study that determined the normative values and prevalence of sarcopenia among a representative sample of 542 community-dwelling adults aged 21–90 years in Singapore. The sampling method has been detailed elsewhere [25]. The final analysis sample consists of 535 participants (308 women, 227 men) who had complete body composition data. All participants provided written informed consent prior to participation and ethics approval was obtained from the National Healthcare Group Domain Specific Review Board (2017/00212).

Anthropometric and body composition measurements

Body weight and height were measured using a digital balance and stadiometer (Seca, GmbH & Co. KG, Hamburg, Germany). BMI was calculated as weight (kg) divided by height (m) squared. Bone mineral density, total body fat percentage, fat mass and appendicular lean mass were measured using dual X-ray absorptiometry (DXA; Discovery WI, Hologic, Inc., Marlborough, USA). FMI and appendicular lean mass index (ALMI) were used as measures of fat mass and muscle mass respectively in the present study to account for height-associated differences in body composition [4,27]. Both FMI and ALMI were calculated as fat mass (kg) and appendicular lean mass (kg) divided by height (m) squared, where fat mass equals to total body fat mass and appendicular lean mass equals to the sum of lean mass in the upper and lower limbs.

Muscle strength and physical function assessment

Handgrip strength (HGS) was used as indication of muscular strength of participants [28] and physical function was measured using usual gait speed (GS) [29]. HGS was assessed using a dynamometer (Jamar Plus+ Dynamometer; Patterson Medical, Evergreen Boulevard, Cedarburg, USA), and the highest of four readings (two trials per arm) recorded. GS was measured using the 6 m GAITRite Walkway (CIR Systems Inc., Sparta. New Jersey, USA) with a 2 m lead in and out phase, and the average speed (three trials) recorded.

Sarcopenia, obesity and sarcopenic obesity assessment

Sarcopenia was assessed using the latest Asian Working Group for Sarcopenia (AWGS) algorithm [4]. Poor physical function was defined as GS <1.0 m/s, low muscle mass as ALMI <7.0 and <5.4 kg/m2, and low muscle strength as HGS <28 and <18 kg for men and women respectively. Sarcopenia was defined as the presence of low muscle mass and poor muscle strength and/or physical performance [4]. Given that there is presently no consensus on the definition of obesity [30], obesity was defined in this study based on the top two quintiles of FMI. The cut-off points adjusted for gender were ≥ 7.63 kg/m2 and ≥ 9.93 kg/m2 for men and women respectively. Participants were categorised into four non-overlapping groups based on presence and absence of sarcopenia and obesity: SO (sarcopenic and obese), sarcopenic (sarcopenic and non-obese), obese (non-sarcopenic and obese) and normal (non-sarcopenic and non-obese).

Cognitive function assessment

Cognitive function was measured using the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS), which encompassed 12 subtests to assess five cognitive domains: immediate memory, visuospatial/constructional abilities, language, attention, and delayed memory [31]. Total scale index and domain-specific index scores for all participants were scored according to the RBANS manual [32], whereby each score was expressed as a standardised score with a mean of 100 and standard deviation of 15. Cognitive impairment was defined using a cut-off index score <80, which corresponds to the 9th percentile [32].

Other measures

Demographic characteristics such as age, gender, education levels, housing type, smoking history and medical history were collected through a questionnaire. Participants also completed the global physical activity questionnaire to report their typical weekly physical activity levels [33].

Statistical analysis

Differences in variables of interest among the four groups defined by sarcopenia and obesity status were examined with one-way analysis of variance for continuous variables and chi-squared test for categorical variables. Multivariate logistic regressions were performed to examine the associations between SO and impairment in global cognition as well as different domains of cognitive function. Odds ratios (OR) and 95% confidence intervals (CI) were estimated to compare the sarcopenic, obese and SO groups with the normal group. In order to examine the relationship between individual components of SO (muscle strength, muscle mass, physical function, and obesity) and cognitive impairment, participants were dichotomised based on each component and separate logistic regressions were conducted to determine the odds of impairment in different domains of cognitive function. Models were adjusted for age groups, gender, physical activity levels, medical history, and years of education. Statistical significance level was set at p < 0.05 and all analyses were performed using R statistical software, version 3.6.3.

Results

The prevalence of SO, sarcopenia, obesity and normal were 7.3%, 18.3%, 33.1% and 41.3% respectively. Table 1 provides the descriptive characteristics of participants across the four groups. The groups were found to differ in prevalence of medical conditions such as diabetes, hypertension, cardiovascular disease, and high cholesterol. HGS and GS were found to differ significantly between groups, whereby the SO group had the lowest HGS and slowest GS. Both the SO and sarcopenic groups were found to be significantly older and had significantly less education as compared to the obese and normal group. The SO group was also found to be significantly less physically active than the other three groups. The prevalence of cognitive impairment in respective domains across the four groups is shown in Fig 1.

Table 1

Descriptive characteristics of participants by sarcopenia and obesity status.

	Sarcopenic Obese (n = 39)	Sarcopenic (n = 98)	Obese (n = 177)	Normal (n = 221)	p-value
Age (y)	75.59 ± 9.00	71.11 ± 14.63	56.12 ± 17.86	51.97 ± 17.95	<0.01*
Age groups					<0.01*
21–39 years	0 (0.0%)	6 (6.1%)	41 (23.2%)	64 (29.0%)
40–59 years	1 (2.6%)	8 (8.2%)	47 (26.6%)	65 (29.4%)
60–74 years	13 (33.3%)	32 (32.7%)	59 (33.3%)	63 (28.5%)
≥75 years	25 (64.1%)	52 (53.0%)	30 (16.9%)	29 (13.1%)
Female (n, %)	25 (64.1%)	52 (53.1%)	100 (56.5%)	131 (59.3%)	0.61
Smoker (n, %)	2 (5.1%)	8 (8.2%)	20 (11.3%)	29 (13.1%)	0.36
Education (y)	6.79 ± 4.93	7.30 ± 5.18	9.95 ± 4.97	10.89 ± 4.55	<0.01*
PA (MET-min/wk)	31.69 ± 27.40	49.75 ± 46.98	75.11 ± 105.26	72.74 ± 73.53	<0.01*
Medical history
Diabetes	9 (23.1%)	27 (27.6%)	24 (13.6%)	20 (9.0%)	< 0.01*
Hypertension	29 (74.4%)	49 (50.0%)	65 (36.7%)	53 (24.0%)	< 0.01*
Cardiovascular disease	4 (10.3%)	12 (12.2%)	7 (4.0%)	9 (4.1%)	0.01*
High cholesterol	27 (69.2%)	44 (44.9%)	73 (41.2%)	58 (26.2%)	<0.01*
Stroke	1 (2.6%)	2 (2.0%)	1 (0.6%)	1 (0.5%)	0.36
Depression	2 (5.1%)	1 (1.0%)	4 (2.3%)	1 (0.5%)	0.11
Body composition
BMI (kg/m2)	25.72 ± 1.78	21.48 ± 2.00	29.21 ± 4.43	22.56 ± 2.55	<0.01*
FMI (kg/m2)	10.68 ± 1.58	7.03 ± 1.55	11.42 ± 2.89	7.08 ± 1.72	<0.01*
ALMI (kg/m2)	5.45 ± 0.71	5.35 ± 0.78	7.04 ± 1.33	6.04 ± 1.13	<0.01*
Physical function
HGS (kg)	20.85 ± 6.10	22.27 ± 6.26	30.50 ± 9.96	30.54 ± 8.98	<0.01*
GS (m/s)	0.84 ± 0.15	0.90 ± 0.15	1.07 ± 0.18	1.15 ± 0.17	<0.01*
Cognitive function
RBANS index score	84.82 ± 14.83	86.41 ± 15.41	93.65 ± 17.19	96.28 ± 15.58	<0.01*
Immediate memory domain	84.69 ± 14.14	88.73 ± 16.31	93.77 ± 16.16	95.10 ± 15.64	<0.01*
Visuospatial domain	89.46 ± 19.38	90.13 ± 15.12	95.98 ± 16.88	99.03 ± 16.43	<0.01*
Language domain	89.56 ± 11.84	88.70 ± 15.30	91.20 ± 15.02	91.28 ± 16.12	0.50
Attention domain	88.05 ± 15.10	91.52 ± 15.23	96.11 ± 19.32	100.47 ± 16.95	<0.01*
Delayed memory domain	90.49 ± 17.01	94.41 ± 16.52	98.55 ± 15.89	100.35 ± 14.30	<0.01*

PA, physical activity levels; BMI, body mass index; FMI, fat mass index; ALMI, appendicular lean mass index; HGS, handgrip strength; GS, gait speed; RBANS, Repeatable Battery for the Assessment of Neuropsychological Status.

*p < 0.05.

All data are presented as mean ± SD or number (%).

Fig 1

Prevalence of cognitive impairment by sarcopenia and obesity status.

Association between cognitive impairment and sarcopenia and obesity status

Table 2 presents the logistic regression analysis results for the association between SO and risk of impairment in different domains of cognitive function. In the unadjusted models, the SO group was found to have significantly higher odds of impairment as compared to the normal group in global cognitive function (OR: 3.75, 95% confidence interval (CI) 1.69–8.13, p < .01), as well as domains of immediate memory (OR: 2.44, 95% CI 1.10–5.21, p = .023), visuospatial/constructional abilities (OR: 4.31, 95% CI 2.00–9.17, p < .01), attention (OR: 2.48, 95% CI 1.05–5.53, p = .031), and delayed memory (OR: 4.18, 95% CI 1.76–9.61, p < .01). Significant association was also found for the sarcopenic group in the unadjusted models for global cognition (OR: 3.00, 95% CI 1.65–5.49, p < .01), and immediate memory (OR: 1.99, 95% CI 1.11–3.54, p = .020). However, the risk of cognitive impairment for the SO and sarcopenic groups were not significant in the adjusted models. The obese group was found to have significantly higher odds of association with impaired global cognition (OR: 1.78, 95% CI 1.03–3.13, p = .041), and attention (OR: 2.24, 95% CI 1.32–3.84, p < .01) in the unadjusted models. After adjusting for confounders, the OR of association between obesity and attention remains significant (OR: 2.08, 95% CI 1.15–3.81, p = .021).

Table 2

Association between sarcopenic obesity and risk of impairment in different domains of cognitive function.

	Sarcopenic Obese (n = 39)		Sarcopenic (n = 98)		Obese (n = 177)		Normal (n = 221)
	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p
Global cognitive function
model 1	3.75 (1.69–8.13)	< .01**	3.00 (1.65–5.49)	< .01**	1.78 (1.03–3.13)	.041*	1.0
model 2	1.55 (0.62–3.77)	.335	1.46 (0.74–2.89)	.274	1.59 (0.88–2.89)	.125	1.0
model 3	1.62 (0.58–4.47)	.350	1.16 (0.53–2.50)	.706	1.55 (0.80–3.02)	.192	1.0
Immediate memory
model 1	2.44 (1.10–5.21)	.023*	1.99 (1.11–3.54)	.020*	1.12 (0.65–1.92)	.673	1.0
model 2	1.38 (0.56–3.27)	.468	1.25 (0.64–2.40)	.504	0.96 (0.54–1.68)	.875	1.0
model 3	1.34 (0.52–3.32)	.534	1.10 (0.55–2.18)	.777	0.87 (0.48–1.55)	.631	1.0
Visuospatial/constructional
model 1	4.31 (2.00–9.17)	< .01**	1.77 (0.93–3.31)	.077	1.30 (0.73–2.29)	.370	1.0
model 2	2.15 (0.87–5.22)	.092	0.85 (0.40–1.74)	.651	1.10 (0.60–2.02)	.757	1.0
model 3	2.09 (0.80–5.44)	.132	0.68 (0.31–1.46)	.334	1.02 (0.53–1.93)	.963	1.0
Language
model 1	0.77 (0.28–1.85)	.592	1.46 (0.82–2.56)	.190	1.05 (0.63–1.73)	.847	1.0
model 2	0.44 (0.14–1.18)	.123	0.98 (0.50–1.86)	.943	0.94 (0.55–1.59)	.822	1.0
model 3	0.40 (0.12–1.12)	.097	0.85 (0.43–1.67)	.644	0.86 (0.49–1.48)	.586	1.0
Attention
model 1	2.48 (1.05–5.53)	.031*	1.73 (0.90–3.27)	.095	2.24 (1.32–3.84)	< .01**	1.0
model 2	1.17 (0.45–2.88)	.737	0.96 (0.46–1.96)	.909	2.05 (1.18–3.60)	.012*	1.0
model 3	1.07 (0.37–2.92)	.904	0.68 (0.30–1.49)	.340	2.05 (1.12–3.82)	.021*	1.0
Delayed memory
model 1	4.18 (1.76–9.61)	< .01**	1.92 (0.92–3.95)	.077	1.35 (0.70–2.64)	.369	1.0
model 2	1.68 (0.63–4.32)	.289	0.79 (0.35–1.77)	.573	1.14 (0.56–2.31)	.713	1.0
model 3	1.67 (0.61–4.45)	.312	0.70 (0.30–1.60)	.405	1.08 (0.52–2.26)	.830	1.0

Model 1: Unadjusted; model 2: Adjusted for age groups, gender, physical activity levels, diabetes, hypertension, cardiovascular disease, high cholesterol, stroke, and depression; model 3: Model 2 and years of education. OR, odds ratio; CI, confidence interval.

*p < .05,

** p < .01.

Table 3 presents the results on the odds of association between individual components of SO and cognitive impairment. After accounting for age groups, gender, physical activity levels, medical history and education, cognitive impairment was found to be significantly associated with low muscle strength, low physical function, and obesity. First, significant association was found between low HGS and impairment in immediate memory (OR: 1.91, 95% CI 1.04–3.49, p = .035). Second, low GS was also significantly associated with greater odds of impairment in the immediate memory (OR: 2.17, 95% CI 1.26–3.72, p < .01. Last, significant relationship was found between high FMI and impairment in attention (OR: 2.06, 95% CI 1.22–3.51, p < .01).

Table 3

Association between components of sarcopenic obesity and risk of impairment in different domains of cognitive function.

	Low ALMI (n = 276)		Low HGS (n = 99)		Low GS (n = 189)		High FMI (n = 216)
	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p
Global cognitive function	0.92 (0.51–1.63)	.768	1.71 (0.88–3.31)	.113	1.67 (0.91–3.03)	.096	1.48 (0.85–2.58)	.164
Immediate memory	0.87 (0.52–1.45)	.600	1.91 (1.04–3.49)	.035*	2.17 (1.26–3.72)	< .01**	0.92 (0.56–1.50)	.742
Visuospatial/constructive	0.80 (0.45–1.38)	.419	1.72 (0.91–3.23)	.094	1.40 (0.77–2.51)	.261	1.40 (0.82–2.39)	.214
Language	0.66 (0.40–1.09)	.106	0.88 (0.45–1.65)	.687	1.07 (0.61–1.83)	.811	0.79 (0.49–1.28)	.349
Attention	0.60 (0.34–1.04)	.069	1.21 (0.62–2.32)	.576	1.64 (0.91–2.92)	.095	2.06 (1.22–3.51)	< .01**
Delayed memory	0.88 (0.47–1.63)	.675	1.57 (0.81–3.02)	.180	1.65 (0.85–3.19)	.138	1.42 (0.79–2.58)	.240

Model adjusted for age groups, gender, physical activity levels, diabetes, hypertension, cardiovascular disease, high cholesterol, stroke, depression, and years of education.

ALMI, appendicular lean mass index; HGS, handgrip strength; GS, gait speed; FMI, fat mass index; OR, odds ratio; CI, confidence interval.

*p < .05,

** p < .01.

We performed additional analyses and found that obesity combined with low HGS or slow GS indeed showed significantly greater odds of cognitive impairment (Tables 4 and 5). After adjusting for confounding variables, co-occurrence of low HGS and high FMI was found to be associated with significantly higher odds of impairment in global cognition (OR: 2.98, 95% CI 1.08–8.19, p = .034), visuospatial/constructional abilities (OR: 3.70, 95% CI 1.45–9.52, p < .01), and attention (OR: 3.11, 95% CI 1.17–8.20, p = .022). Similarly, significant association was found between impairment in the attention domain and co-occurrence of slow gait speed and high FMI (OR: 3.02, 95% CI 1.41–6.51, p < .01).

Table 4

Association between handgrip strength, obesity, and risk of impairment in different domains of cognitive function.

	Low HGS and Obese (n = 35)		Low HGS (n = 64)		Obese (n = 181)		Normal (n = 255)
	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p
Global cognitive function	2.98 (1.08–8.19)	.034*	1.58 (0.69–3.57)	.275	1.42 (0.75–2.69)	.279	1.0
Immediate memory	1.95 (0.78–4.76)	.145	1.76 (0.84–3.64)	.128	0.89 (0.50–1.57)	.695	1.0
Visuospatial/constructive	3.70 (1.45–9.52)	< .01**	1.12 (0.49–2.49)	.782	1.07 (0.57–1.99)	.837	1.0
Language	0.46 (0.14–1.31)	.166	1.08 (0.51–2.26)	.830	0.90 (0.53–1.53)	.708	1.0
Attention	3.11 (1.17–8.20)	.022*	1.03 (0.43–2.39)	.940	1.87 (1.03–3.41)	.039*	1.0
Delayed memory	2.37 (0.88–6.24)	.083	1.58 (0.67–3.66)	.284	1.44 (0.71–2.96)	.313	1.0

Model adjusted for age groups, gender, physical activity levels, diabetes, hypertension, cardiovascular disease, high cholesterol, stroke, depression, and years of education.

HGS, handgrip strength; OR, odds ratio; CI, confidence interval.

*p < .05,

** p < .01.

Table 5

Association between gait speed, obesity, and risk of impairment in different domains of cognitive function.

	Low GS and Obese (n = 88)		Low GS (n = 101)		Obese (n = 128)		Normal (n = 218)
	OR (95% CI)	p	OR (95% CI)	p	OR (95% CI)	p
Global cognitive function	2.20 (0.99–4.91)	.054	1.94 (0.89–4.21)	.092	1.74 (0.81–3.73)	.151	1.0
Immediate memory	1.87 (0.92–3.78)	.083	2.29 (1.17–4.47)	.015*	0.90 (0.43–1.79)	.761	1.0
Visuospatial/constructive	1.86 (0.87–3.98)	.107	1.26 (0.59–2.65)	.547	1.26 (0.59–2.63)	.544	1.0
Language	0.86 (0.41–1.75)	.681	1.10 (0.56–2.12)	.767	0.80 (0.42–1.48)	.477	1.0
Attention	3.02 (1.41–6.51)	< .01**	1.54 (0.70–3.34)	.279	2.01 (0.98–4.11)	.056	1.0
Delayed memory	2.17 (0.93–5.10)	.073	1.59 (0.69–3.68)	.276	1.36 (0.55–3.28)	.491	1.0

Model adjusted for age groups, gender, physical activity levels, diabetes, hypertension, cardiovascular disease, high cholesterol, stroke, depression, and years of education.

GS, gait speed; OR, odds ratio; CI, confidence interval.

*p < .05,

** p < .01.

Discussion

This study examined the association between SO and cognitive impairment among healthy community-dwelling adults. We found that sarcopenia with and without obesity were not associated with global or domain-specific cognitive impairment after adjusting for potential confounders. However, obesity on its own was significantly associated with greater odds of impairment in the attention domain. Furthermore, we identified that low muscle strength, low physical function and high fat mass were individually associated with impaired cognitive function.

Sarcopenia alone [6,7] and concurrent with obesity [9,21,22] have been purported to be associated with cognitive impairment. While we did observe that sarcopenia and SO were related to cognitive impairment in the unadjusted models, these associations were not significant after accounting for age, gender, and education. Increasing age and lower education levels are well established strong risk factors for Alzheimer’s disease [34], and strongly confounded the observed higher prevalence of cognitive impairment in SO and sarcopenic individuals, who were significantly much older and less educated as compared to the other groups. The diagnosis of sarcopenia requires the presence of low muscle mass as the core criterion [4]. The salient lack of independent association between low ALMI and all domains of cognitive function critically explains the insignificant association between sarcopenia and cognitive impairment.

On the other hand, we found that other functional components of sarcopenia, i.e. low HGS and GS, were individually associated with cognitive impairment even after accounting for confounding variables. These findings corroborate previous studies that suggested muscle quality is more relevant to cognitive function than muscle quantity [8,35-38]. Such results are also consistent with previous evidence of association between GS, HGS, and cognitive impairment [39,40]. Although the underlying mechanisms remain unclear, mutual pathological factors such as chronic inflammation, oxidative stress, hormonal changes, and insulin resistance have been postulated to explain the association between poor physical function and cognitive decline [41,42]. Neurological mechanisms are also plausible given that both HGS and GS performance involve multiple brain regions [43,44]. The present study findings revealed that low HGS and slow GS were associated with impairment in immediate memory domain. Future studies are warranted to elucidate the associated pathophysiological pathways.

Interestingly, we found that obesity on its own was associated with significantly greater risk of cognitive impairment. While obesity has been posited to have links with reduced cognitive function [45,46], there is criticism that the relationship between adiposity and cognitive impairment is obscured by the use of BMI to define obesity [47], since it is a measure of excess weight instead of body fat [48]. The present study is confirmatory in demonstrating that greater FMI was associated with cognitive impairment after accounting for confounding variables. Previous studies have reported poorer executive function among obese individuals [49,50]. Our results extend current body of knowledge by showing that greater adiposity was specifically associated with deficit in attention domain. This association is plausibly explained through systemic inflammation and insulin resistance [19]. Increased adipose tissues have been reported to increase secretion of proinflammatory cytokines [51] and impair insulin sensitivity [52], in which both conditions were associated with cognitive impairment among older adults [53,54].

SO was postulated to have stronger detrimental effects on cognitive function than either sarcopenia or obesity alone [9,21]. It has been demonstrated to affect multiple cognitive domains among community-dwelling adults [21] and type 2 diabetes patients [55]. In this study, we found that SO was not associated with cognitive impairment in adjusted models. This was due to the lack of independent effects of ALMI (as the core criterion of sarcopenia) on cognitive function. Our results indicate that in the absence of low ALMI, the co-occurrence of obesity with either low muscle strength or poor physical performance (as alternative definitions of sarcopenic obesity) were significantly associated with increased odds of global and domain-specific cognitive impairment. Thus, our observations strongly suggest that the impacts of sarcopenia and SO on cognitive outcome is stronger if based on muscle strength and physical performance parameters without the requisite core criterion of low ALMI.

Insulin resistance-induced endothelial dysfunction may mediate the relationship between sarcopenic obesity and cognitive function [9,56]. Beyond its catabolic effect on skeletal muscle tissues, insulin resistance has been suggested to correlate with poorer muscle function independently [57]. Thus, obese individuals with greater insulin resistance may be at greater risk of decline in muscle strength and physical performance. Since insulin resistance is not measured in present study, further research is required to examine the suggested mechanism behind the relationship between SO and cognitive impairment.

The strengths of the present study are the recruitment of a representative sample across different adult age groups and the use of a gold standard instrument to measure body composition. However, we have to acknowledge some limitations in the present study. First, as the study sample consisted of relatively healthy community-dwelling adults, caution is advised in generalising the results to individuals with more severe physical and cognitive impairments. Second, due to its cross-sectional nature, our study is not able to determine the temporality and direction of the associations identified. Future longitudinal studies are necessary to establish the specific relationship between components of SO and cognitive impairment.

In conclusion, we found no association between sarcopenic obesity based on AWGS’s definition of sarcopenia and cognitive impairment after adjusting for confounders. Low lean mass was notably not associated with cognitive impairment, but low muscle strength, low physical performance and obesity defined by high FMI were independently associated. The co-occurrence of obesity with either low muscle strength or poor physical performance was associated with exacerbated odds of observing cognitive impairment.

Dataset.

(CSV)

Click here for additional data file.