Sarcopenia and decline in appendicular skeletal muscle mass are associated with hypoperfusion in key hubs of central autonomic network on 3DSRT in older adults with progression of normal cognition to Alzheimer's disease

Abstract

Aim

Although sarcopenia is common in patients with Alzheimer's disease (AD), the neural substrates involved remain unclear. We investigated the relationship between sarcopenia, as well as its definition components, and regional cerebral blood flow (rCBF) in older adults with progression of normal cognition to AD.

Methods

^99m Tc‐ethyl‐cysteinate‐dimer single‐photon emission computed tomography was carried out in 95 older adults with progression of normal cognition to AD (40 men and 55 women, mean ± SD age 80.9 ± 6.8 years). The associations of rCBF determined by 3‐D stereotactic region of interest template software, with sarcopenia and its definition components, slower gait speed, weaker grip strength, and decline in appendicular skeletal muscle mass index (ASMI) were analyzed.

Results

Logistic regression analysis adjusted by age, sex, mini‐mental state examination score and education showed that sarcopenia as well as ASMI less than the cut‐off (men 7.0 kg/m², women 5.7 kg/m²) were associated with significantly reduced rCBF in the key hub of the central autonomic network, including the insula, anterior cingulate cortex, subcallosal area, rectal gyrus, hypothalamus, amygdala and caudate head. Sarcopenia and ASMI decline were associated with hypoperfusion in the aforementioned cortical hubs of the central autonomic network in men, but with hypoperfusion of the hypothalamus in women. Linear regression analysis showed significant correlations of ASMI/cut‐off with rCBF in the bilateral medial frontal cortex, as well as rCBF in the aforementioned key hubs.

Conclusions

Hypoperfusion in key hubs of central autonomic network is implicated in the emergence of sarcopenia, probably through ASMI decline in vulnerable older adults. Geriatr Gerontol Int 2023; 23: 16–24.

SPM8 results from group comparison of regional cerebral blood flow (rCBF). Subjects with sarcopenia (n = 56) showed significant reduction of rCBF in the bilateral anterior cingulate cortex and left insula compared to those without sarcopenia (n = 39).

Introduction

Sarcopenia, defined as an age‐dependent pathological decline in muscle mass and function ¹ occurring late in life, is often associated with impaired autonomic function manifested as lower parasympathetic activity‐associated dysfunctional heart rate modulation ² and orthostatic hypotension ³ in community‐dwelling older adults. In contrast, sarcopenia more often emerges in older adults with mild cognitive impairment and/or Alzheimer's disease (AD) compared with cognitively robust older adults. ⁴ Furthermore, community‐dwelling older adults with more severe sarcopenia have a higher risk of development of incident AD. ⁵ The natural disease process of AD involves not only progressive deterioration in cognitive function, but also dysfunction of autonomic modulation to autonomic stimuli/tasks. ⁶ , ⁷ , ⁸ Impairment of the central autonomic network (CAN) could be causatively involved in these autonomic impairments, as neurodegenerative processes have an impact on areas that are important to CAN with varying degrees in AD patients. ⁹ , ¹⁰

CAN has been proposed by Benarroch ¹¹ as the neural system through which the brain integrates emotional, cognitive, neuroendocrine, visceromotor and sensory information into homeostatic autonomic responses essential for survival. Recent meta‐analyses of human neuroimaging studies have shown significant roles of forebrain structures, including the thalamus, insula, anterior cingulate cortex, medial frontal cortex, hypothalamus, amygdala and hippocampus in the CAN, ¹² , ¹³ besides brainstem structures. ¹⁴ Tight connections of the caudate, putamen and globus pallidus with these key hubs in response to stimuli/tasks for CAN were also reported. ¹³ , ¹⁵ Among these key hubs of CAN, cortical areas (insula, anterior cingulate cortex and medial frontal cortex) are implicated in high‐order autonomic control, projecting directly to other CAN areas, and the hypothalamus is implicated in neuroendocrine and stress responsivity. ¹⁶

However, the details of this three‐way integration of sarcopenia, AD and impaired CAN function remain unclear. The present study aimed to elucidate the relationship between sarcopenia and regional cerebral blood flow (rCBF) using ^99m Tc‐ethyl‐cysteinate‐dimer (ECD) single‐photon emission computed tomography in older adults with progression of normal cognition (NC) to AD, who are most vulnerable to rCBF change with functional declines in cognition and CAN. ⁹

Methods

Study population

Cognitive function was assessed by the Mini‐Mental State Examination (MMSE). The present study included 95 older adults with progression of NC to AD (40 men and 55 women, mean ± SD age 80.9 ± 6.8 years, range 66–93 years) who attended our outpatient clinic, including 29 older adults with NC (MMSE 28–30), 23 with mild cognitive impairment (MMSE 24–27) and 43 with AD (MMSE 11–23). Probable AD was diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition. ¹⁷ All these participants underwent general neurological and physical examinations, brain evaluation by magnetic resonance imaging and MR angiography, and laboratory tests to exclude other potential causes of dementia. None of the participants had received sex hormone replacement therapy or had used sex hormone for treatment. Individuals with a history of dementia with stroke, frontotemporal dementia, dementia with Lewy bodies, Parkinson's disease or major depression, and individuals with carotid stenosis or intracranial large artery stenosis on MR angiography were excluded. Outpatients using a wheelchair were also excluded.

Sarcopenia

Appendicular skeletal muscle mass index (ASMI) was measured by a multifrequency bioelectrical impedance analysis using the InBody720 (BioSpace, Seoul, Korea) with cut‐off values of <7.0 kg/m² for men and <5.7 kg/m² for women, based on the Asian Working Group for Sarcopenia criteria. ¹ Usual gait speed was assessed with a Walk‐way system (MW1000; Anima, Tokyo, Japan), with a cut‐off of <1.0 m/s. ¹ Dominant grip strength was assessed with a Smedley‐type handheld dynamometer with cut‐off values of <28 kg for men and <18 kg for women. ¹ Sarcopenia was defined by the Asian Working Group for Sarcopenia as the combination of ASMI less than the cut‐off plus gait speed less than the cut‐off and/or grip strength less than the cutoff. ¹

Image acquisition

All older adults underwent brain perfusion ^99m Tc‐ECD single‐photon emission computed tomography. Participants were injected with 600 MBq ^99m Tc‐ECD after a 5‐min rest, and further rested for another 5 min in a quiet room with their eyes closed. Single‐photon emission computed tomography projection data were obtained five times for 4 min each starting 10 min after intravenous injection of ^99m Tc‐ECD, with a Symbia Evo Excel (Siemens Healthcare, Tokyo, Japan). The photo peak was centered on 141 keV in 90 projections with 360° rotation (128 × 128 matrix format).

Univariate image analysis

A voxel‐by‐voxel group study was then carried out using statistical parametric mapping 8 software (SPM8; http://www.fil.ion.ucl.ac.uk/spm, University College, London, UK) and MATLAB Compiler Runtime version R2013a (Mathworks, Cambridge, UK). Images were initially converted from the Digital Imaging and Communications in Medicine (DICOM) to Neuroimaging Informatics Technology Initiative (NifTI) and transferred to SPM8. The data were then standardized with the Montreal Neurological Institute atlas. Standardized data were then smoothed by a Gaussian filter (FWHM 12 mm). Differences in rCBF between the presence and absence of sarcopenia and its definition components were assessed using a general linear model (two‐sample t‐test) in SPM8 (height threshold P < 0.001).

Assessment of rCBF

3‐D stereotactic region of interest template (3DSRT v. 3.6; Fujifilm RI Pharma, Tokyo, Japan) image analysis was used for rCBF quantification. This program measures rCBF in 46 paired segments (paired segments of 41 regions [v. 3.0] ¹⁸ plus paired segments of hypothalamus, quadrigemini, substantia nigra, nucleus ruber and pons [v. 3.6]) with an anatomical standardization engine, and covers almost the whole cerebral and cerebellar gray matter in only a few minutes. Mean rCBF through each segment was expressed as the percentage of whole brain blood flow. The paired segments of interest in the present study were thalamus, insula, anterior cingulate cortex, subcallosal area, rectal gyrus, medial frontal cortex, hypothalamus, amygdala, hippocampus, caudate head, putamen and globus pallidus.

Statistical methods

The χ²‐test was used for categorical variables, and the Mann–Whitney U‐test for the distribution of ordinal variables to compare the two groups. Logistic regression analysis was used to assess the associations of each rCBF with sarcopenia and its definition components, slower gait speed, weaker grip strength and ASMI decline, adjusted by age, sex, MMSE score and education. Linear regression analysis was carried out using ASMI/cut‐off as the outcome variable, and each rCBF as the predictor variable adjusted by age, sex, MMSE score and education. The ability to predict rCBF less than the median for ASMI/cut‐off was assessed as the area under the curve (AUC) on receiver operating characteristic (ROC) analysis. The optimal ROC curve point for the prediction of rCBF less than the median was defined as the ASMI/cut‐off value that maximized sensitivity plus specificity. P < 0.05 was adopted as significant. Data were analyzed using SPSS Windows (v. 29.0; SPSS, Chicago, IL, USA).

Ethical considerations

The protocol for this study was approved by the clinical research ethics committee of Kanazawa Medical University Hospital (approval No. I361). All participants gave written informed consent.

Results

Hypoperfusion area

SPM8 analysis showed that participants with sarcopenia (n = 56) had hypoperfusion in the bilateral anterior cingulate cortex and left insula compared with those without sarcopenia (n = 39; Fig. 1a), and that participants with ASMI less than the cut‐off (n = 58) had hypoperfusion in the bilateral anterior cingulate cortex and insula compared with those with ASMI equal to or greater than the cut‐off (n = 37; Fig. 1b).

Figure 1.

SPM8 results from group comparison of regional cerebral blood flow (rCBF). Participants with sarcopenia (n = 56) showed a significant reduction of regional cerebral blood flow in the bilateral anterior cingulate cortex (0 38 –4, x y z; kE 109) and left insula (−36 16 6, x y z; kE 65) compared with those without sarcopenia (n = 39) (a). Participants with appendicular muscle mass index less than the cut‐off (men: 7.0 kg/m², women: 5.7 kg/m²; n = 58) showed a significant reduction of regional cerebral blood flow in the bilateral anterior cingulate cortex (0 38 –4, x y z; kE 167), left insula (−36 16 6, x y z; kE 67) and right insula (35 17 6, x y z; kE 104) compared with those without (n = 37) (b). The height threshold was <0.001, corrected for multiple comparisons. SPM8‐derived x, y and z; spatial coordinates in Montreal Neurological Institute space as well as kE; cluster level.

Baseline characteristics and rCBF

Further detailed analysis of the association of CAN hubs with sarcopenia and its definition components was carried out using 3DSRT. Table 1 summarizes the baseline characteristics and rCBF in key hubs of CAN compared between two groups according to sex, cognitive function, and the presence versus absence of sarcopenia and its definition components. The incidence of sarcopenia was significantly higher in participants with mild cognitive impairment and those with AD compared with those with NC. Women had significantly higher mean rCBF in the bilateral hypothalamus and caudate head compared with men.

Table 1.

Baseline characteristics and regional cerebral blood flow in groups according to sex, cognitive function, sarcopenia as defined by the Asian Working Group for Sarcopenia and its definition components

Characteristics	Sex		Cognitive function			AWGS sarcopenia		Gait speed <1.0 m/s		Grip strength <cut‐off 1		ASMI <cut‐off
	Men†	Women	NC†	MCI††	AD	(−)†	(+)	(−)†	(+)	(−)†	(+)	(−)†	(+)
	n = 40	n = 55	n = 29	n = 23	n = 43	n = 39	n = 56	n = 18	n = 77	n = 34	n = 61	n = 37	n = 58
Demographics
Age (years)	80.9 ± 6.1	80.9 ± 7.1	79.6 ± 5.8	81.2 ± 7.8	81.6 ± 6.5#	78.0 ± 5.9	82.9 ± 6.4**	75.1 ± 5.6	82.2 ± 6.1***	76.9 ± 6.3	83.1 ± 5.8***	78.8 ± 5.5	82.2 ± 7.0*
Women, n (%)			16 (55.2%)	14 (60.9%)	25 (58.1%)	21 (53.9%)	34 (60.7%)	13 (72.2%)	42 (54.6%)#	19 (55.9%)	36 (59.0%)	20 (54.1%)	35 (60.3%)
Education (years)	11.6 ± 2.5	10.6 ± 2.6*	11.1 ± 2.6	11.3 ± 2.8	10.9 ± 2.5	11.4 ± 2.5	10.8 ± 2.7	11.7 ± 2.8	10.9 ± 2.5	11.8 ± 2.5	10.6 ± 2.6*	11.2 ± 2.4	11.0 ± 2.7
MMSE (/30)	23.7 ± 5.1	23.2 ± 5.4	29.0 ± 0.9	25.6 ± 1.0***	18.5 ± 3.4***+++	24.1 ± 5.9	22.9 ± 4.7#	25.4 ± 4.4	22.9 ± 5.3#	24.1 ± 5.4	23.0 ± 5.1#	24.1 ± 6.0	22.9 ± 4.7#
BMI (kg/m2)	22.5 ± 3.6	22.3 ± 3.6	22.8 ± 2.7	21.2 ± 3.3#	22.8 ± 2.7$	23.8 ± 3.7	21.5 ± 3.2**	21.6 ± 4.0	22.6 ± 3.5#	23.1 ± 4.0	22.5 ± 3.3	24.3 ± 3.2	21.2 ± 3.3***
AWGS sarcopenia, n (%)	22 (55.0%)	33 (60.0%)	11 (37.9%)	17 (73.9%)*	27 (62.8%)*
Gait speed <1.0 m/s, n (%)	35 (87.5%)	42 (76.4%)#	20 (69.0%)	21 (91.3%)#	36 (83.7%)#
Grip strength <cut‐off, n (%)	25 (62.5%)	36 (65.5%)	16 (55.2%)	16 (69.6%)	29 (67.4%)
ASMI <cut‐off, n (%)	23 (57.5%)	35 (63.6%)	12 (41.4%)	18 (78.3%)*	28 (65.1%)*
rCBF (‰)
Thalamus R	37.7 ± 3.0	38.7 ± 3.4#	38.4 ± 3.7	38.8 ± 2.9	37.9 ± 3.2$	38.5 ± 3.3	38.1 ± 3.2	39.7 ± 3.0	37.9 ± 3.2*	38.3 ± 4.2	38.4 ± 2.7	38.7 ± 3.4	38.0 ± 3.1
Thalamus L	36.6 ± 3.3	38.0 ± 3.4#	37.5 ± 3.5	38.4 ± 3.1	36.7 ± 3.4$	37.9 ± 3.4	37.0 ± 3.4#	39.2 ± 2.4	36.9 ± 3.5**	37.8 ± 3.6	37.1 ± 3.3	38.0 ± 3.6	37.0 ± 3.3#
Insula R	42.4 ± 1.7	42.8 ± 1.6	42.4 ± 1.6	42.7 ± 1.5	42.7 ± 1.7	43.0 ± 1.3	42.3 ± 1.8*	43.3 ± 1.2	42.4 ± 1.7*	42.6 ± 1.6	42.6 ± 1.6	43.1 ± 1.3	42.3 ± 1.7*
Insula L	42.1 ± 1.8	42.0 ± 1.7	42.2 ± 1.8	42.1 ± 1.8	41.9 ± 1.7	42.7 ± 1.4	41.5 ± 1.8**	42.8 ± 1.5	41.9 ± 1.8#	42.0 ± 1.3	42.1 ± 2.0	42.8 ± 1.3	41.6 ± 1.8**
Anterior cingulate cortex R	35.2 ± 2.6	35.0 ± 2.5	36.2 ± 2.7	35.1 ± 2.5#	34.3 ± 2.2**$	36.2 ± 2.7	34.3 ± 2.1***	36.1 ± 2.4	34.8 ± 2.5#	35.2 ± 2.5	35.0 ± 2.5	36.3 ± 2.8	34.3 ± 2.0***
Anterior cingulate cortex L	34.6 ± 2.9	34.5 ± 2.4	35.5 ± 2.7	34.8 ± 2.2	33.7 ± 2.5**+	35.5 ± 2.9	33.8 ± 2.1***	35.9 ± 2.4	34.2 ± 2.5*	34.7 ± 2.6	34.4 ± 2.6	35.6 ± 3.0	33.9 ± 2.1**
Subcallosal area R	39.1 ± 2.5	39.8 ± 3.3#	39.8 ± 3.1	40.3 ± 2.7	38.9 ± 3.0*$	40.5 ± 2.3	38.8 ± 3.2**	40.8 ± 2.2	39.2 ± 3.0*	39.6 ± 2.5	39.5 ± 3.2	40.6 ± 2.3	38.8 ± 3.1**
Subcallosal area L	40.2 ± 2.6	39.8 ± 3.0	40.4 ± 2.8	40.4 ± 2.2	39.5 ± 3.1#$	40.7 ± 2.5	39.5 ± 3.0*	41.0 ± 2.4	39.7 ± 2.9#	40.0 ± 2.6	40.0 ± 3.0	40.9 ± 2.5	39.4 ± 2.9**
Rectal gyrus R	37.2 ± 3.0	37.1 ± 2.9	37.5 ± 3.3	37.6 ± 2.7	36.9 ± 2.8$	37.9 ± 2.7	36.7 ± 3.0#	37.4 ± 2.7	37.1 ± 2.9	37.0 ± 2.4	37.2 ± 3.2	38.0 ± 2.8	36.6 ± 2.8*
Rectal gyrus L	37.5 ± 2.9	37.1 ± 3.1	38.0 ± 3.4	37.3 ± 2.7	36.8 ± 2.9#	38.1 ± 2.8	36.7 ± 3.1*	37.9 ± 3.0	37.1 ± 3.0	37.2 ± 2.7	37.3 ± 3.2	38.3 ± 2.9	36.6 ± 2.9*
Medial frontal cortex R	43.0 ± 2.4	43.6 ± 2.5	43.6 ± 2.1	42.1 ± 2.5*	43.8 ± 2.5+	43.7 ± 2.7	43.1 ± 2.2	43.0 ± 2.1	43.4 ± 2.5	43.5 ± 2.7	43.3 ± 2.3	43.7 ± 2.8	43.1 ± 2.2
Medial frontal cortex L	43.1 ± 2.6	43.5 ± 2.7	43.7 ± 2.0	42.3 ± 2.6#	43.7 ± 2.9$	43.7 ± 2.8	43.1 ± 2.4	43.1 ± 2.5	43.4 ± 2.6	43.6 ± 2.9	43.2 ± 2.4	43.7 ± 2.9	43.2 ± 2.4
Hypothalamus R	27.0 ± 3.1	28.9 ± 3.4*	27.5 ± 3.5	29.1 ± 3.6	28.0 ± 3.2	28.9 ± 3.2	27.6 ± 3.5*	28.8 ± 3.3	28.0 ± 3.4#	27.7 ± 3.5	28.3 ± 3.4	29.2 ± 3.0	27.5 ± 3.5**
Hypothalamus L	26.9 ± 3.0	28.7 ± 3.7**	27.3 ± 3.3	29.1 ± 3.7#	27.8 ± 3.5	28.6 ± 3.4	27.5 ± 3.6*	29.3 ± 2.7	27.6 ± 3.6*	27.9 ± 3.2	28.0 ± 3.7	28.9 ± 3.5	27.3 ± 3.5*
Amygdala R	29.4 ± 3.3	30.1 ± 3.2	30.7 ± 2.6	30.6 ± 3.4	28.8 ± 3.4*$	30.3 ± 2.9	29.4 ± 3.4#	30.8 ± 2.4	29.6 ± 3.4#	30.0 ± 2.9	29.7 ± 3.5	30.5 ± 3.1	29.4 ± 3.3#
Amygdala L	28.9 ± 3.0	29.4 ± 3.0	30.1 ± 2.8	29.4 ± 3.5	28.5 ± 2.8*	30.1 ± 2.6	28.5 ± 3.1**	30.2 ± 2.2	29.0 ± 3.1#	29.2 ± 2.5	29.2 ± 3.3	30.4 ± 2.8	28.4 ± 2.9**
Hippocampus R	33.8 ± 2.3	33.6 ± 2.4	34.2 ± 2.3	34.4 ± 2.0	33.5 ± 2.5#$	33.9 ± 2.1	33.5 ± 2.6	34.5 ± 2.0	33.5 ± 2.4#	33.9 ± 2.4	33.6 ± 2.4	34.0 ± 2.2	33.5 ± 2.5
Hippocampus L	32.9 ± 2.3	33.0 ± 2.5	33.7 ± 2.4	33.1 ± 2.3	32.4 ± 2.3*	33.3 ± 2.0	32.7 ± 2.6*	33.6 ± 1.9	32.8 ± 2.5	32.9 ± 2.4	33.0 ± 2.4	33.5 ± 2.3	32.6 ± 2.4*
Caudate head R	28.6 ± 4.1	30.3 ± 3.6*	30.3 ± 3.6	30.5 ± 3.9	28.7 ± 3.9#$	30.4 ± 4.1	29.0 ± 3.6*	31.3 ± 3.4	29.2 ± 3.9*	29.4 ± 3.9	29.7 ± 3.9	30.7 ± 4.1	28.9 ± 3.6**
Caudate head L	27.7 ± 3.9	29.7 ± 3.5**	29.6 ± 3.6	30.0 ± 3.6	27.8 ± 3.8#$	29.9 ± 4.0	28.2 ± 3.5**	31.3 ± 3.1	28.3 ± 3.7**	29.1 ± 3.8	28.7 ± 3.8	30.0 ± 4.1	28.2 ± 3.4**
Putamen R	46.0 ± 2.1	46.2 ± 2.2	45.6 ± 2.1	46.1 ± 1.6	46.4 ± 2.4#	46.4 ± 2.1	45.9 ± 2.2	46.8 ± 2.4	46.0 ± 2.1	45.8 ± 2.4	46.3 ± 2.0#	46.5 ± 2.1	45.8 ± 2.2#
Putamen L	45.4 ± 2.0	45.4 ± 2.0	44.8 ± 2.0	45.6 ± 1.3#	45.7 ± 2.2#	45.6 ± 1.8	45.2 ± 2.1	45.7 ± 2.1	45.3 ± 2.0	45.2 ± 2.2	45.5 ± 1.9	45.7 ± 1.7	45.2 ± 2.2#
Globus pallidus R	42.6 ± 1.9	43.3 ± 2.2#	42.2 ± 2.2	43.3 ± 1.7#	43.4 ± 2.2*	43.1 ± 2.1	43.0 ± 2.2	43.4 ± 2.1	42.9 ± 2.1	42.8 ± 2.3	43.1 ± 2.0	43.2 ± 2.1	42.9 ± 2.2
Globus pallidus L	42.8 ± 2.0	43.6 ± 2.1#	42.2 ± 2.4	43.6 ± 1.2**	43.7 ± 2.1**	43.5 ± 2.3	43.1 ± 2.0	43.5 ± 2.4	43.2 ± 2.1	43.4 ± 2.2	43.1 ± 2.0	43.7 ± 2.2	42.9 ± 2.1#

Note: Demographic results are expressed as mean ± SD or n (%). Results of rCBF are shown as permillage (‰) of whole brain blood flow. χ² analysis or Mann–Whitney U‐analysis was used. ^# P < 0.20, *P < 0.05, **P < 0.01 and ***P < 0.001 versus ^†group for each category, and ^$ P < 0.20, ⁺ P < 0.05, ⁺⁺ P < 0.01 and ⁺⁺⁺ P < 0.001 versus ^††group. Cut‐off values for grip strength were 28 kg for men and 18 kg for women, and those for appendicular skeletal muscle mass index (ASMI) were 7.0 kg/m² for men and 5.7 kg/m² for women.

Abbreviations: AD, Alzheimer‐type dementia; AWGS, Asian Working Group for Sarcopenia; BMI, body mass index; MCI, mild cognitive impairment; MMSE, Mini‐Mental State Examination; NC, normal cognition.

¹ Right‐dominant grip strength was observed in 54 participants and left‐dominant grip strength in 31 participants.

Logistic regression analysis

Logistic regression analysis adjusted by age, sex, MMSE score and education showed that sarcopenia, as defined by the Asian Working Group for Sarcopenia, in total participants was associated with significantly reduced rCBF in the bilateral anterior cingulate cortex, rectal gyrus, hypothalamus and caudate head; left insula and amygdala; and right subcallosal area (Fig. 2a). In addition, ASMI less than the cut‐off in total participants was associated with significantly reduced rCBF in the bilateral insula, anterior cingulate cortex, subcallosal area, rectal gyrus, hypothalamus and caudate head, and left amygdala and globus pallidus (Fig. 2b). Logistic regression analysis adjusted by age, MMSE score and education showed that sarcopenia (Fig. 2a) and ASMI less than the cut‐off (Fig. 2b) were associated with significantly reduced rCBF in the bilateral insula, anterior cingulate cortex, subcallosal area and rectal gyrus, and in the left amygdala in men, but not in women, whereas they were associated with significantly reduced rCBF in the left hypothalamus in women, but not in men. Logistic regression analysis showed that slower gait speed (<1.0 m/s) was associated with significantly (P = 0.028) reduced rCBF in the left caudate head with an odds ratio of 0.79 (95% confidence interval 0.64–0.98) in total participants, but not with rCBF in any other key hubs of CAN, and that weaker grip strength was not significantly associated with rCBF in any of these key hubs. Logistic regression analysis adjusted by age and sex showed that progression of NC to AD was associated with significantly reduced rCBF in limbic key hubs, including the bilateral anterior cingulate cortex and amygdala, and in the left hippocampus in total participants, and with significantly reduced rCBF in the bilateral amygdala and left hippocampus in participants with sarcopenia (Figure S1).

Figure 2.

Logistic regression odds ratios for regional cerebral blood flow in (a) participants with sarcopenia compared with those without, and in (b) participants with appendicular muscle mass index (ASMI) less than the cut‐off compared with those without. Error bars represent 95% confidence intervals (CIs). Results are adjusted by age, sex, Mini‐Mental State Examination score and education in total participants (n = 95), and adjusted by age, Mini‐Mental State Examination score and education in men (n = 40) and in women (n = 55). *P < 0.05, **P < 0.01 and ***P < 0.001.

Linear regression analysis

Figure 3 shows regression coefficients (B) of linear regression analysis using ASMI/cut‐off as the outcome variable and rCBF as the predictor variable, adjusted by age, sex, MMSE score and education in total participants, and by age, MMSE score and education in men and in women, respectively. Linear regression analysis showed that ASMI/cut‐off was significantly associated with rCBF in the bilateral medial frontal cortex besides the key hubs, with a significant relation to ASMI less than the cut‐off in logistic regression analysis, including the left insula and bilateral anterior cingulate cortex, hypothalamus, and caudate head in total participants (Fig. 3). A sex difference was again observed in the sites of CAN hubs; ASMI/cut‐off was significantly associated with rCBF in the bilateral insula, anterior cingulate cortex, subcallosal area and medial frontal cortex, and left caudate head in men, but not in women, whereas it was significantly associated with rCBF in the bilateral hypothalamus in women, but not in men (Fig. 3).

Figure 3.

Linear regression coefficients (B) of regional cerebral blood flow on appendicular muscle mass index/cut‐off. Error bars represent 95% confidence intervals. Results are corrected for age, sex, Mini‐Mental State Examination score and education in total participants (n = 95), and by age, Mini‐Mental State Examination score and education in men (n = 40) and in women (n = 55). *P < 0.05, **P < 0.01 and ***P < 0.001.

ROC curve analysis

The ability of ASMI/cut‐off to predict rCBF less than the median in the left insula, anterior cingulate cortex, hypothalamus and amygdala in the total participants, in men, and in women was compared using ROC curve analysis (Fig. 4). The AUC was 66% (P = 0.008) in the left insula, 70% (P = 0.001) in the left anterior cingulate cortex, 66% (P = 0.008) in the left hypothalamus and 66% (P = 0.008) in the left amygdala in total participants. The optimal ROC curve points of ASMI/cut‐off chosen as the values maximizing sensitivity plus specificity for the prediction of rCBF less than the median in the left insula, anterior cingulate cortex, hypothalamus and amygdala were 1.00, 0.97, 0.97 and 0.99, respectively (Fig. 4). Although AUCs were not significant (P > 0.05) when analyzed separately in men and women, the optimal ROC curve points of ASMI/cut‐off for rCBF less than the median in the left insula, anterior cingulate cortex, amygdala and hypothalamus ranged from 0.97 to 1.00 (Fig. 4).

Figure 4.

Predictive ability of appendicular muscle mass index (ASMI) for reduction of regional cerebral blood flow. Receiver operating characteristic (ROC) curves were created to predict regional cerebral blood flow less than the median in the left insula, anterior cingulate cortex (ACC), hypothalamus and amygdala on ASMI/cut‐off (men 7.0 kg/m², women 5.7 kg/m²), in total participants (n = 95), in men (n = 40) and in women (n = 55). Arrows indicate optimal ROC curve points chosen as values maximizing sensitivity plus specificity. AUC, area under the receiver operating characteristic curve

Discussion

We previously reported involvement of frontal lobe shrinkage in sarcopenia in older adults with progression of NC to AD. ¹⁹ SPM8 analysis visualized brain regions with hypoperfusion relating to sarcopenia and ASMI less than the cut‐off (Fig. 1). Our 3DSRT analysis further detailed the attenuated forebrain structures reported to be key hubs of CAN, ¹² , ¹³ including the insula, anterior cingulate cortex, subcallosal area, rectal gyrus, hypothalamus and amygdala, with close relationships to sarcopenia and ASMI decline (Fig. 2). Atrophy of limbic key hubs, including the amygdala and hippocampus (Figure S1), might explain the integration of sarcopenia and AD. ⁴ , ⁵ The association of caudate atrophy with slower gait speed shown in the present study has also been reported in older adults. ²⁰ Linear regression analysis in the present study also showed that ASMI/cut‐off was significantly associated with rCBF in the medial frontal cortex besides rCBF in the above key hubs of CAN in the total participants (Fig. 3). ROC curve analysis (Fig. 4) showed significant AUCs when comparing ASMI/cut‐off with rCBF less than the median of the left insula, anterior cingulate cortex, hypothalamus and amygdala, suggesting that all these CAN hubs exert coordinated roles in the maintenance of ASMI. The optimal ROC curve points of ASMI/cut‐offs predicting rCBF less than the median in these CAN hubs were very near to 1.00 in the total participants, suggesting the cut‐off values of ASMI (7.0 kg/m² in men and 5.7 kg/m² in women) ¹ would indicate the almost median point of rCBF reduction in these CAN hubs.

The results of the present logistic regression analysis showed a sex difference in the sites of CAN hubs associated with sarcopenia and ASMI decline. Sarcopenia (Fig. 2a) and ASMI less than the cut‐off (Fig. 2b) were associated with significantly reduced rCBF in telencephalic structures, including the bilateral insula, anterior cingulate cortex, subcallosal area and rectal gyrus, and in the left amygdala in men, but not in women. Linear regression analysis also showed that ASMI/cut‐off was significantly associated with rCBF in the bilateral medial frontal cortex besides the aforementioned key hubs of CAN in men, but not in women (Fig. 3). Although the mechanism(s) of this association in men is unclear, a possible explanation is that attenuation of these high‐order autonomic cortices could have a causal role in inadequate reinnervation of skeletal muscle fibers. Sarcopenia is characterized by progressive loss of motoneurons largely due to inadequate reinnervation of muscle fibers by the remaining motoneurons. ²¹ Sympathetic nerves in muscle interact with both blood vessels and muscle fibers, playing roles in regulation of skeletal muscle homeostasis. ²² A recent study suggested that a decline in reinnervation of the skeletal muscle by both motor and sympathetic axons might cause the emergence of sarcopenia. ²³ Furthermore, the burst incidence of resting postganglionic muscle sympathetic nerve activity has been shown to largely correlate with the strength of connectivity for the high‐order key hubs of CAN, including the left insula and right anterior cingulate cortex. ²⁴ Furthermore, previous studies analyzing cardiovascular autonomic response by functional imaging have shown higher activation of the insular and anterior cingulate cortex in men compared with women. ²⁵ , ²⁶

On the contrary, sarcopenia and ASMI decline were associated with significantly reduced rCBF in the left hypothalamus in women, but not in men (Fig. 2). Linear regression analysis also showed that ASMI/cut‐off was significantly associated with rCBF in the bilateral hypothalamus in women, but not in men (Fig. 3). The precise reason(s) for this association in women is unclear. However, growth hormone (GH)‐releasing hormone produced by the hypothalamus mediates both pituitary GH and hepatic insulin‐like growth factor‐1 (IGF‐1), which are known to be anabolic factors for protein synthesis of muscle. ²⁷ The declines in circulating GH and IGF‐1 after menopause in women are likely to accelerate the loss of muscle. ²⁷ Low‐dose injections of recombinant human GH‐releasing hormone for 3 months resulted in significant elevation of circulating concentrations of GH and IGF‐1, and decreases in stair‐climbing and walking times in older women. ²⁸ Furthermore, lower serum IGF‐1 was independently associated with higher odds for sarcopenia in community‐dwelling older women, but not in men. ²⁹ These outcomes suggest that attenuation of the hypothalamo–pituitary–somatotropic axes would be critical in homeostasis of skeletal muscle mass in older women.

Laterality of CAN hubs in association with sarcopenia and ASMI decline was also observed in the present study, as reduced rCBF in the left insula was related to sarcopenia in total participants, and reduced rCBF in the left amygdala was related to sarcopenia and ASMI decline in total participants and in men (Fig. 2). A recent functional magnetic resonance imaging study in healthy controls and patients with behavioral variant frontotemporal dementia, a disease characterized by baseline autonomic dysfunction, showed that left lateralized CAN hubs, including the left anterior cingulate and left insular cortices, facilitate parasympathetic outflow. ³⁰ The precise mechanisms by which hypoperfusion in CAN hubs and sympathetic and/or parasympathetic dysregulation emerge as sarcopenia and ASMI decline should be elucidated in the future.

Several limitations of the present study must be considered. First, this was a single‐center study, and a multicenter study is required for further evaluation. Second, hypoperfusion of precise nuclei or areas in each of 46 paired segments was not elucidated due to the resolution limit of 3DSRT, and should be clarified using functional imaging in the future. Third, a possible relationship between functional decline in these key hubs of CAN and sarcopenia/ASMI decline could also be clarified in community‐dwelling older adults and in individuals with other neurodegenerative disorders in the future.

In conclusion, the present study showed that sarcopenia, as well as ASMI decline, was associated with hypoperfusion in key hubs of CAN in older adults with progression of NC to AD. The precise mechanism(s) of these associations should be clarified in the future.

Disclosure Statement

The authors declare no conflict of interest.

Supporting information

Figure S1. Logistic regression odds ratios for regional cerebral blood flow in participants with Alzheimer‐type dementia (AD) compared with those with normal cognition (NC), in participants with mild cognitive impairment (MCI) compared with those with NC, and in participants with AD compared with those with MCI, in total participants (n = 95), in participants with sarcopenia (n = 56) and in participants without sarcopenia (n = 39). Error bars represent 95% confidence intervals (CIs). Results are adjusted by age and sex. *P < 0.05 and **P < 0.01.

Demura T, Okuno T, Miwa T, et al. Sarcopenia and decline in appendicular skeletal muscle mass are associated with hypoperfusion in key hubs of central autonomic network on 3DSRT in older adults with progression of normal cognition to Alzheimer's disease. Geriatr. Gerontol. Int. 2023;23:16–24. 10.1111/ggi.14515

Data Availability Statement

Data are available on request because of privacy/ethical restrictions.