The importance of neuromuscular rate of force development for physical function in aging and common neurodegenerative disorders – a systematic review

Abstract

We systematically reviewed existing literature regarding lower extremity neuromuscular rate of force development (RFD), maximal muscle strength (Fmax), and physical function in neurodegenerative populations, and to what extent these outcomes are affected and/or associated. Following PRISMA guidelines, 4 databases (Pubmed, Embase, SPORTDiscus, Web of Science) were searched. Across aging, Parkinson Disease (PD), Alzheimer’s Disease (AD), Multiple Sclerosis (MS), or Stroke, included studies should report (Part 1) deficits in lower extremity RFD, Fmax, and physical function (~ individuals having inferior vs. superior physical function), and/or (Part 2) associations between RFD (or Fmax) and physical function. A total of N=32 studies (n=1087 participants) were included. Part 1: deficits in RFD (-31%, mean; N=22) were comparable to deficits in physical function (-26%; N=7), yet both deficits exceeded that of Fmax (-21%; N=20). Part 2: associations between RFD and physical function (r²=0.13, mean; N=16) were comparable to associations between Fmax and physical function (r²=0.15; N=12). Lower extremity RFD is (1) particularly sensitive (i.e. adapts earlier and/or more extensively) towards neurodegeneration, and more so than Fmax, and (2) of importance for physical function but apparently not superior to Fmax. RFD could serve as a useful indicator/biomarker of changes in neuromuscular function elicited by neurodegeneration.

Introduction

Neurological disorders of the central nervous system (CNS), more specifically neurodegenerative disorders, affect millions of people worldwide[1-4]. These disorders afflict people across all ages, yet particularly at advanced age, with the number of neurodegenerative patients estimated to increase considerably in the decades to come[2]. The deleterious changes of the central nervous system that accompany neurodegenerative disorders – a phenomenon also observed with aging[5] – altogether constitute major causes of disability[2-4,6,7].

Some of the most common and prevalent neurodegenerative disorders comprise Parkinson Disease (PD), Alzheimer’s Disease (AD), Multiple Sclerosis (MS), and Stroke[2-4,8]. These disorders induce marked structural and functional changes of the central nervous system (e.g. loss of neurons, glial cells, and myelin along with reductions in motor unit recruitment and firing frequency responsible for muscle activation/control)[5,9-11]. PD is characterized by loss of dopaminergic neurons within the Substantia Nigra[12,13]. AD is characterized by amyloid plaques and neurofibrillary tangles causing macroscopic atrophy from synaptic and neuronal loss[14]. MS is characterized by demyelination and axonal loss within the CNS[15]. Stroke differs as it is characterized by an acute episode of focal dysfunction of the brain, retina or spinal cord[16], yet it has been argued to be accompanied by secondary neurodegeneration over time, thus resembling some of the processes observed with PD, AD, and MS[11]. With aging, neurodegenerative processes also occur and resemble those observed with PD, AD, and MS – albeit less pronounced – comprising loss of neurons, glial cells, myelin etc[5,7,13].

A common consequence of the structural and functional CNS changes described above is that neuromuscular function (i.e. the nerve-muscle interaction responsible for motor function) deteriorates, initiating a cascade of deleterious events. Specifically, the outlined neurodegenerative populations are accompanied by impairments in neuromuscular function (i.e. ability to generate and control muscle force), often preferentially affecting the lower extremities[17-20], that eventually causes limitations in physical function (i.e. ability to perform activities of daily living such as walking) and ultimately disability[17,21-30]. To counteract this, it is of paramount importance to identify and understand relevant and modifiable neuromuscular predictors of physical function / disability.

The most examined feature of neuromuscular function is maximal muscle strength, defined as the ability of a muscle or muscle group to generate maximal force (or torque) during a voluntary contraction (termed force max; Fmax), with numerous studies having reported moderate-to-strong associations between lower extremity Fmax and physical function (e.g. walking) in the outlined neurodegenerative populations[31-37]. In contrast, the rate by which muscle strength can be developed, defined as the ability to increase force (or torque) as rapidly as possible during a voluntary contraction (termed rate of force development; RFD)[38], may contribute independently to physical performance[39], particularly in tasks requiring fast body movements. Previous studies have even argued RFD to be more sensitive to detect acute and/or chronic adaptations in neuromuscular function compared to Fmax[38,40,41]. The proposed reason for this is that RFD is particularly reliant on a well-functioning CNS[38,42]. In aging and neurodegenerative disorders, it is thus likely that RFD is preferentially prone to changes due to the substantial neurodegeneration observed in these populations. Also, in line with the aforementioned information emphasizing the importance of Fmax, some studies (yet fewer in number than with Fmax) have reported moderate-to-strong associations between lower extremity RFD and physical function (e.g. between knee extension RFD and walking) in aging[39,43] as well as in neurodegenerative disorders such as PD[44], MS[33], and Stroke[45]. Nonetheless, we are unaware of any systematic reviews evaluating whether RFD (compared to Fmax) is preferentially affected and associated with physical function in the outlined neurodegenerative populations. Therefore, the aim of this study was to carry out a systematic review summarizing existing evidence from studies investigating lower extremity RFD alongside physical function in aging and common neurodegenerative disorders (PD, AD, MS and Stroke). Hopefully, such information can help advance our understanding of RFD and its implications for physical function, and to clarify whether RFD can serve as a useful indicator (or biomarker) associated with neurodegeneration.

Design and Methods

Literature search

This review is based on a systematic literature search of Pubmed, Embase, SPORTDiscus, and Web of Science that was performed to identify scientific articles investigating lower extremity RFD alongside physical function in populations preferentially undergoing neurodegeneration, i.e. aging, PD, AD, MS, and Stroke. The literature search included articles published before April 24^th 2020. Using the Boolean operators ‘and’ and ‘or’, five search strategies concerning aging, Parkinson Disease, Alzheimer’s Disease, Multiple Sclerosis, and Stroke were performed combined with the following search terms: ‘rate of force development’, ‘rate of torque development’, ‘rate of strength development’, ‘strength development rate’ and ‘explosive muscle strength’. In addition to a free text search, each neurodegenerative condition was also identified using Mesh-terms or Emtree-terms in Pubmed and Embase, respectively. Filters were applied on Pubmed and Embase to encompass study populations only comprised of humans. Regarding aging, the filter was limited to ages 65+ years. For the exact search strategies and terms used in the various databases, please see Supplementary Table 1. Finally, reference lists of the included studies were checked for relevant articles. This systematic review is composed in accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA)[46]. Furthermore, Cochrane’s Covidence served as a tool for screening and identification of duplicates.

Supplementary Table 1.

Search Strategies. Table showing the exact search strategies in different databases.

Database	Number	Search terms
Pubmed
Aging	465	(((((((aging[MeSH Terms]) OR “aging”) OR “ageing”) OR “aged”) OR old*) OR elder*)) AND (((((“rate of force development”) OR “rate of torque development”) OR “rate of strength development”) OR “strength development rate”) OR “explosive muscle strength”) Filters: Humans; Aged: 65+ years
Parkinson Disease	17	(((parkinson disease[MeSH Terms]) OR parkinson*)) AND (((((“rate of force development”) OR “rate of torque development”) OR “rate of strength development”) OR “strength development rate”) OR “explosive muscle strength”) Filters: Humans
Multiple Sclerosis	7	(((multiple sclerosis[MeSH Terms]) OR “multiple sclerosis”)) AND (((((“rate of force development”) OR “rate of torque development”) OR “rate of strength development”) OR “strength development rate”) OR “explosive muscle strength”) Filters: Humans
Stroke	63	(((stroke[MeSH Terms]) OR stroke*)) AND (((((“rate of force development”) OR “rate of torque development”) OR “rate of strength development”) OR “strength development rate”) OR “explosive muscle strength”) Filters: Humans
Alzheimer’s Disease	9	((((Dementia[MeSH Terms]) OR “dementia”) OR alzheimer*)) AND (((((“rate of force development”) OR “rate of torque development”) OR “rate of strength development”) OR “strength development rate”) OR “explosive muscle strength”) Filters: Humans
Embase
Aging	229	(‘aging’/exp OR ‘aging’ OR ‘ageing’ OR ‘aged’ OR old* OR elder*) AND (‘rate of force development’ OR ‘rate of torque development’ OR ‘rate of strength development’ OR ‘strength development rate’ OR ‘explosive muscle strength’) AND [aged]/lim AND [humans]/lim
Parkinson Disease	16	(‘parkinson disease’/exp OR parkinson*) AND (‘rate of force development’ OR ‘rate of torque development’ OR ‘rate of strength development’ OR ‘strength development rate’ OR ‘explosive muscle strength’) AND [humans]/lim
Multiple Sclerosis	7	(‘multiple sclerosis’/exp OR ‘multiple sclerosis’) AND (‘rate of force development’ OR ‘rate of torque development’ OR ‘rate of strength development’ OR ‘strength development rate’ OR ‘explosive muscle strength’) AND [humans]/lim
Stroke	25	(‘cerebrovascular accident’/exp OR stroke*) AND (‘rate of force development’ OR ‘rate of torque development’ OR ‘rate of strength development’ OR ‘strength development rate’ OR ‘explosive muscle strength’) AND [humans]/lim
Alzheimer’s Disease	2	(‘dementia’/exp OR ‘dementia’ OR alzheimer*) AND (‘rate of force development’ OR ‘rate of torque development’ OR ‘rate of strength development’ OR ‘strength development rate’ OR ‘explosive muscle strength’) AND [humans]/lim
SPORTDiscus
Aging	157	( “aging” OR “ageing” OR “aged” OR old* OR elder* ) AND ( “rate of force development” OR “rate of torque development” OR “rate of strength development” OR “strength development rate” OR “explosive muscle strength” )
Parkinson Disease	6	parkinson* AND ( ”rate of force development” OR ”rate of torque development” OR ”rate of strength development” OR ”strength development rate” OR ”explosive muscle strength” )
Multiple Sclerosis	1	”multiple sclerosis” AND ( ”rate of force development” OR ”rate of torque development” OR ”rate of strength development” OR ”strength development rate” OR ”explosive muscle strength” )
Stroke	25	stroke* AND ( “rate of force development” OR “rate of torque development” OR “rate of strength development” OR “strength development rate” OR “explosive muscle strength” )
Alzheimer’s Disease	0	( “dementia” OR alzheimer* ) AND ( “rate of force development” OR “rate of torque development” OR “rate of strength development” OR “strength development rate” OR “explosive muscle strength” )
Web of Science
Aging	304	TOPIC: (“aging” OR “ageing” OR “aged” OR old* OR elder*) AND TOPIC: (“rate of force development” OR “rate of torque development” OR “rate of strength development” OR “strength development rate” OR “explosive muscle strength”)
Parkinson Disease	19	TOPIC: (parkinson*) AND TOPIC: (“rate of force development” OR “rate of torque development” OR “rate of strength development” OR “strength development rate” OR “explosive muscle strength”)
Multiple Sclerosis	8	TOPIC: (“multiple sclerosis”) AND TOPIC: (“rate of force development” OR “rate of torque development” OR “rate of strength development” OR “strength development rate” OR “explosive muscle strength”)
Stroke	38	TOPIC: (stroke*) AND TOPIC: (“rate of force development” OR “rate of torque development” OR “rate of strength development” OR “strength development rate” OR “explosive muscle strength”)
Alzheimer’s Disease	1	TOPIC: (“dementia” OR alzheimer*) AND TOPIC: (“rate of force development” OR “rate of torque development” OR “rate of strength development” OR “strength development rate” OR “explosive muscle strength”)

Inclusion and exclusion criteria

Only studies including older individuals (mean age ≥65 years) or subjects with definite PD, AD, MS or Stroke were included in this review, with additional inclusion criteria being 1) available in English, Danish or German; 2) human studies; and 3) assessment of lower extremity RFD and physical function (see definitions below). The exclusion criteria were 1) reviews; 2) case reports including ≤4 participants; 3) abstracts only; and 4) interventions not reporting baseline data. One investigator (SDL) performed the initial assessment of eligibility identifying potentially relevant studies to include, while another investigator (LGH) performed assessment of eligibility of those potentially relevant studies. Consensus was subsequently reached between the two investigators.

Definition of outcomes and coding of studies

The terms Fmax (defined as the ability of a muscle or muscle group to generate maximal force or torque during a voluntary contraction) and RFD (defined as the ability to increase force or torque as rapidly as possible during a voluntary contraction) were both used throughout this review, derived from a force-time (or torque-time) curve. As RFD can be derived and expressed in many different ways (e.g. as slopes from the onset of contraction to specific time points or as specific regions of the rising phase of a muscle contraction)[38], all measures were included. Furthermore, we defined early phase RFD (comprising RFD derived from the onset of contraction to time points ≤100 ms) and late phase RFD (comprising RFD derived from the onset of contraction to time points 150-300 ms). While both early and late phase RFD rely on a well-functioning CNS, this has been argued to be especially pronounced in early phase RFD[38,42], providing a rationale for examining this in-depth in the outlined neurodegenerative populations.

The following classification of lower extremity physical function was used: Short walk tests, comprising walking tests ≤10 m; Long walk tests, comprising 400 m walk test, 6 minute walk test (6MWT), 2 minute walk test (2MWT); Chair rise tests, comprising sit-to-stand tests (STS), chair rise/stand; Time-up-and-go tests (TUG), comprising traditional TUG as well as Expanded Timed Get-up-and-go (ETGUG), 8-Feet Up-and-Go; Short Physical Performance Battery (SPPB); Stair climb tests; and Balance tests, comprising static/dynamic sway/posturagraphy analysis (also including perturbations), balance score, stability index, Flamingo Balance Test, Fullerton Advanced Balance scale etc. History of falls were also used, representing a dichotomous classification of individuals having inferior vs. superior lower extremity physical function, respectively.

Data extraction and analysis

Characteristics of the participants (number of participants, age, gender, time since diagnosis (patients only), disease stage (patients only)) along with mean or median of the study outcomes (i.e. measures of lower extremity Fmax, RFD, Function) were extracted. Data was included according to two types of analyses: Part 1) parallel observations of baseline deficits in lower extremity Fmax, RFD, and physical function (along with associations between these deficits, subsequently calculated by the present study investigators) and Part 2) reported associations between lower extremity RFD and physical function (denoted RFD-Function) as well as Fmax and physical function (denoted Fmax-Function). Of note, we allowed that Part 1 could contain deficit data from studies comparing individuals having inferior vs. superior physical function, such as patients vs. healthy controls, older vs. young adults, and older fallers vs. non-fallers etc. Regarding Part 2, associations were considered weak if r²<0.16, moderate if 0.16≤r²<0.49, strong if 0.49≤r²<0.81, and very strong if r²≥0.81[47].

Recollecting that RFD can be derived and expressed in several different ways, all reported measures were included yet summarized to represent one lower extremity RFD value from each study. Also, if studies reported data on more than one muscle group or action (such as flexion and extension), data on Fmax and RFD was summarized to represent one value, respectively, from each study. The same approach was done for early and late phase RFD, respectively. Table 1 and Supplementary Table 2 contain detailed characteristics of the identified studies including all extracted outcomes of Fmax, RFD, and physical function (including specific muscle actions as well as the approach used to derive/calculate RFD). WebPlotDigitizer (https://apps.automeris.io/wpd/) was used to extract numerical data in studies where only graphical plots were published.

Table 1.

Deficits (Part 1).

	Study			Deficit (%)			Muscle actions	Assessment details
				Fmax	RFD	Function
Older non-fallers vs. fallers	Bento et al. 2010#	Non-fallers Fallers	n=13 n=18	-17.6	-17.2		Dorsi+Plantar flex, Hip ext+flex+abd+add, Knee ext+flex	RFD20-80%
	Crozara et al. 2013	Non-fallers Fallers	n=22 n=21	-11.1	-10.3		Dorsi+Plantar flex, Knee ext+flex	RFD50ms
	Kamo et al. 2019	Non-fallers Fallers	n=34 n=88	-2.4	-13.8		Knee ext	RFD200ms
	LaRoche et al. 2010	Non-fallers Fallers	n=12 n=11	-18.5	-12.8		Dorsi+Plantar flex, Knee ext+flex	RFD200ms
	Morcelli et al. 2016a,b	Non-fallers Fallers	n=24 n=20	-10.7	-24.1		Hip ext+flex+abd+add	RFD50ms, RFD100ms, RFD150ms, RFD200ms
	Palmer et al. 2015#	Non-fallers Fallers	n=9 n=6	-35.3	-37.3		Hip ext	RFD50ms, RFD100-200ms
	Pijnappels et al. 2008	Non-fallers Fallers	n=10 n=7	-26.1	-39.3		Knee ext, Leg press, Plantar flex	RFD100ms
	Weighted mean 95% CI [lower: upper]	Participants Studies	n=295 N=7	-10.8 [-12.2 : -9.4]	-17.8 [-25.5 : -10.0]
Older lower vs. higher functioning	Clark et al. 2013#	Fast walkers Slow walkers	n=12 n=8		-40.7	-14.7	Plantar flex	RFDmax Long walkmax, Short walkusual+max, SPPB
	LaRoche et al. 2011	Normal strength Low strength	n=11 n= 13	-23.7	-28.0	-11.0	Dorsi+Plantar flex Knee ext+flex	RFD200ms Chair rise, Short walkusual, SPPB
	Palmer et al. 2016#	Higher function Lower function	n=9 n=6	-19.2	-36.1		Knee ext+flex	RFD50ms, RFD200ms
	Weighted mean 95% CI [lower: upper]	Participants Studies	n=59 N=3	-22.0 [-31.9 : -12.0]	-34.4 [-40.8 : -27.9]	-12.7 [-13.7 : -11.7]
Older vs. younger adults	Crozara et al. 2013	Young Old	n=18 n= 43	-39.0	-59.3		Dorsi+Plantar flex Knee ext+flex	RFD50ms
	Inacio et al. 2019#	Young Old	n=15 n=15	-34.4	-49.8		Hip abd+add	RFDpeak
	Izquierdo et al. 1999	Young Old	n=12 n=10	-46.6	-64.9		Squat	RFDpeak
	Mackey et al. 2006	Young Old	n=25 n=25	-4.3	-15.6		Plantar flex	RFD0-85%
	Morcelli et al. 2016a,b	Young Old	n=18 n=44	-46.5	-62.6		Hip ext+flex+abd+add	RFD50ms, RFD100ms, RFD150ms, RFD200ms
	Palmer et al. 2017#	Young Old	n=11 n=11	-17.6	-24.3		Hip ext	RFD50ms, RFD100ms, RFD150ms, RFD200ms
	Sundstrup et al. 2010	Young Old	n=49 n=18	-21.7	-21.8		Knee ext	RFD100ms, RFD200ms
	Unhjem et al. 2019#	Young Old	n=9 n=32	-4.8	-31.2	-23.4	Leg press (dyn)	RFD30ms, RFD50ms, RFD100ms, RFD150ms, RFD200ms Chair rise, Short walkusual, Stair climb
	Weighted mean 95% CI [lower: upper]	Participants Studies	n=355 N=8	-27.0 [-41.6 : -12.3]	-40.8 [-59.0 : -22.5]	-23.4
PD patients vs. healthy controls	Malling et al. 2016	Healthy controls PD patients	n=17 n=13		-25.4	-33.0	STS (up+down)	RFD0-70% Chair rise
	Noorvee et al. 2006#	Healthy controls PD patients	n=12 n=12	-6.1	-26.2	-13.2	Knee ext	RFD200ms Short walkusual
	Pääsuke et al. 2002#	Healthy controls PD patients	n=12 n=14	-7.3	-35.5	-49.3	Knee ext (unilat.), STS (up)	RFDmax Chair rise
	Pääsuke et al. 2004#	Healthy controls PD patients	n=16 n=12	-19.5	-32.6	-31.7	Knee ext (bilat.), STS (up)	RFDmax Chair rise
	Weighted mean 95% CI [lower : upper]	Participants Studies	n=108 N=4	-11.3 [-19.1 : -3.5]	-30.4 [-36.1 : -24.7]	-30.7 [-47.3 : -14.1]
All	Weighted mean 95% CI [lower: upper]	Participants Studies	n=817 N=22	-20.5 [-29.4 : -11.7]	-30.6 [-40.8 : -20.4]	-25.9 [-36.1 : -15.6]

STS: sit to stand. SPPB: short physical performance battery. #: Fmax and RFD outcomes that were initially reported as absolute values, but subsequently normalised by body mass reported by the study. In studies reporting data from more than one muscle action, mean deficit values of Fmax and RFD (based on a mean RFD value in studies reporting data on more than one RFD measure), respectively, were calculated and presented. In studies reporting data on more than one measure of physical function, a mean deficit value of Function was calculated and presented.

Supplementary Table 2.

Description of all identified studies.

Study	Study type	Groups (n) (Women:Men) Age	Times since diagnosis/ Disease stage	MVC	RFD	Functional Outcome(s)
Cross-sectional studies
Aging
Altubasi et al. 2015 Elderly	Cross-sectional	Elderly (n=21) (13:8) 71.3±4.6	NA	KE: 163.6±57.7 Nm	RFDmaxKE: 108.7±64.3 Nm/s	Stairclimbing: 4.7±0.9 s Ramp up walk: 2.2±0.3 s TUG: 7.1±1.1 s 4 m walking time: 3.1±0.6 s
Data presented as mean±SD		Equipment: Biodex Dynamometer
Bento et al. 2010 Elderly with no fall history vs. one fall vs. ≥2 falls	Cross-sectional	≥2 falls (n=10) (10:0) 67.8±8.8	NA	HAd: 90.96±52.66 Nm HAb: 104.25±59.58 Nm HF: 70.21±44.68 Nm HE: 129.55±66.19 Nm KF: 28.11±10.32 Nm KE: 70.27±29.1 Nm DF: 20.59±19.59 Nm PF: 23.31±13.22 Nm	RFD(20-80%MVC):  HAd: 0.85±0.63 Nm/s  HAb: 0.72±0.58 Nm/s  HF: 0.76±0.55 Nm/s  HE: 1.49±1.09 Nm/s  KF: 0.23±0.13 Nm/s  KE: 0.50±0.33 Nm/s  DF: 0.09±0.09 Nm/s  PF: 0.21±0.17 Nm/s	≥2 falls within the last 12 mo
		1 fall (n=8) (8:0) 66.0±4.9	NA	HAd: 74.47±44.68 Nm HAb: 87.23±28.19 Nm HF: 61.70±8.51 Nm HE: 112.23±51.6 Nm KF: 27.53±10.61 Nm KE: 65.68±20.36 Nm DF: 15.51±6.35 Nm PF: 21.95±7.71 Nm	RFD(20-80%MVC):  HAd: 0.80±0.59 Nm/s  HAb: 0.76±0.28 Nm/s  HF: 0.72±0.19 Nm/s  HE: 1.21±0.74 Nm/s  KF: 0.25±0.09 Nm/s  KE: 0.59±0.19 Nm/s  DF: 0.14±0.03 Nm/s  PF: 0.24±0.09 Nm/s	1 fall within the last 12 mo
		Non-fallers (n=13) (13:0) 67.6±7.5	NA	HAd: 110.11±45.21 Nm HAb: 106.38±48.94 Nm HF: 70.21±36.7 Nm HE: 157.45±111.7 Nm KF: 44.46±27.53 Nm KE: 76.58±43.31 Nm DF: 18.98±13.31 Nm PF: 27.03±16.53 Nm	RFD(20-80%MVC):  HAd: 1.02±0.63 Nm/s  HAb: 0.89±0.5 Nm/s  HF: 0.74±0.57 Nm/s  HE: 1.55±1.0 Nm/s  KF: 0.44±0.29 Nm/s*  KE: 0.71±0.47 Nm/s  DF: 0.13±0.11 Nm/s  PF: 0.24±0.20 Nm/s	No fall history
Data presented as mean±SD Strength data extracted from figure		*Larger than the two other groups of fallers		Equipment: Load cell
Clark et al. 2013 Elderly faster vs. slower walkers	Cross-sectional	Faster (n=12) (6:6) 70.8±4.5	NA	NR	VGRFDPF, dyn: 3218±442 N/s*	Δusual-max speed > 0.6 m/s 10 m usual speed: 1.37±0.15 m/s 10 m max speed: 2.17±0.20 m/s* 400 m usual speed: 1.31±0.15 m/s* SPPB: 11.8±0.6* BBT: 54.7±1.3
		Slower (n=8) (4:4) 71.4±5.0	NA	NR	VGRFDPF, dyn: 2010±1112 N/s	Δusual-max speed: > 0.6 m/s 10 m usual speed: 1.24±0.15 m/s 10 m max speed: 1.76±0.16 m/s 400 m usual speed: 1.1±0.11 m/s SPPB: 10.1±1.2 BBT: 55.1±1.1
Data presented as mean±SD		*Different from Slower		Equipment: Bertec force plate
Crockett et al. 2013 Elderly	Cross-sectional	Elderly (n=29) (17:12) 67.3±5.5	NA	KE Con90°/s: 144.7±42.9 Nm Con180°/s: 70.6±21.5 Nm Ecc: 183.1±54.9 Nm	RFDmaxKE:  Con90°/s: 393.1±159.7 Nm/s  Con180°/s: 396.7±135.2 Nm/s  Ecc: 287.0±92.2 Nm/s	30s STS: 14.0±3.6
Data presented as mean±SD	Equipment: Humac Norm dynamometer
Crozara et al. 2013 Elderly fallers and non-fallers vs. healthy young females	Cross-sectional	Elderly fallers (n=21) (21:0) 69.62±7.16	NA	KE: 1.32±0.27 Nm/kg KF: 0.54±0.07 Nm/kg PF: 0.74±0.29 Nm/kg DF: 0.34±0.08 Nm/kg	RFD50KE: 2.51± 1.23 Nm/s/kg  KF: 0.94±0.47 Nm/s/kg  PF: 1.18±0.65 Nm/s/kg  DF: 0.83±1.20 Nm/s/kg	Falls within the last year
		Elderly non-fallers (n=22) (22:0) 66.14±6.1	NA	KE: 1.52±0.28 Nm/kg* KF: 0.64±0.18 Nm/kg* PF: 0.76±0.21 Nm/kg DF: 0.39±0.08 Nm/kg	RFD50KE: 2.87±1.16 Nm/s/kg  KF: 1.16±0.58 Nm/s/kg  PF: 1.29±0.5 Nm/s/kg  DF: 0.84±0.41 Nm/s/kg	No falls within the last year
		Young (n=18) (18:0) 21.79±2.12	NA	KE: 2.61±0.45 Nm/kg*§ KF: 1.26±0.26 Nm/kg*§ PF: 1.36±0.29 Nm/kg*§ DF: 0.48±0.11 Nm/kg*§	RFD50KE: 7.23±3.15 Nm/s/kg*§  KF: 3.63±1.38 Nm/s/kg*§  PF: 2.90±1.55 Nm/s/kg*§  DF: 1.55±0.44 Nm/s/kg*§	__
Data presented as mean±SD Strength data extracted from figure		*Different from fallers. §Different from non-fallers	Equipment: Biodex dynamometer
Hester et al. 2020 >Elderly	Cross-sectional	Elderly (n=26) (19:7) 73.73±4.9	NA	HG: 29.04±10.6 kg	VGRFDPF,dyn: NR	5-Chair rise: NR 30 s Chair rise: NR Preferred walking speed: NR Maximal walking speed: NR TUG
Data presented as mean±SD	Equipment: Force plate
Inacio et al. 2019 Young vs. old	Cross-sectional	Elderly (n=15) (6:9) 71.3±0.9	NA	HAb: 0.039±0.002 Nm/m×kg* HAd: 0.061±0.001 Nm/m×kg* VGRFLateral stepping:  50% BW:0.105±0.004 N/m×kg*  65% BW:0.116±0.003 N/m×kg*  80% BW: 0.139±0.005 N/m×kg	RFDpeakHAb: 73.3±6.3 Nm/s* HAd: 145.3±12.5 Nm/s VGRFDmaxLateral stepping:  50% BW: 1777.3±101.4 N/s*  65% BW: 1598.3±106.5 N/s*  80% BW: 1245.7±86.8 N/s*	Lateral balance perturbations: Incidens of lateral stepping: 50% BW: 73.3±8.0% 65% BW: 65.3±7.5% 80% BW: 43.4±8.6%
		Young (n=15) (8:7) 29.1±1.1	NA	HAb: 0.063±0.001 Nm/m×kg HAd: 0.088±0.002 Nm/m×kg VGRFLateral stepping: 50%  BW: 0.06±0.002 N/m×kg 65%  BW: 0.09±0.002 N/m×kg 80%  BW: 0.14±0.04 N/m×kg	RFDpeakHAb: 213.7±20.5 Nm/s HAd: 179.5±16.8 Nm/s VGRFDmaxLateral stepping:  50% BW: 642.1±38.6 N/s  65% BW: 619.5±32.6 N/s  80% BW: 732.6±37.1 N/s	Lateral balance perturbations: Incidens of lateral stepping: 50% BW: 92.5±3.2%* 65% BW: 94.0±2.9%* 80% BW: 85.4±4.0%*
Data presented as mean±SE	*Different from Young	Equipment: Biodex dynamometer & AMTI force platform
Izquierdo et al. 1999 Young vs. middle-aged vs. old	Cross-sectional	Young (~20 y) (n=12) (0:12) 21±1	NA	Squat, static: 1381±81 N	VGRFDpeaksquat, static: 8474.03±616.8 N/s	Balance:  Time of transition: 1.72±0.27 s  Time inside center: 81.92±2.39 s  Straightness of trajectory: 75±2%  Distance center of pressure: 5814±387 mm*  Balance area: 4926±215 mm2*
		Middle-aged (~40 y) (n=10) (0:10) 40±2	NA	Squat, static: 1039±92 N§	VGRFDpeaksquat, statis: 8246.75±1071.4 N/s	Balance:  Time of transition: 2.01±0.6 s  Time inside center: 75.38±2.61 s  Straightness of trajectory: 74±3%  Distance center of pressure:  10707±2372 mm  Balance area: 5546±857 mm2
		Elderly (~70 y) (n=10) (0:10) 71±5	NA	Squat, static: 747±63 N*§	VGRFDpeaksquat, statis: 3019.80±389.6 N/s*§	Balance:  Time of transition: 2.51±0.71 s*§  Time inside center: 55.77±6.54 s*§  Straightness of trajectory: 76±5%  Distance center of pressure:  9463±2079 mm  Balance area: 6305±627 mm2
Characteristics of subjects presented as mean±SD Strength and functional data presented as mean±SE RFD data extracted from figure		*Different from Middle-aged §Different from Young	Equipment: Dinascan force platform
Kamo et al. 2019 Elderly with no fall history vs. one fall vs. ≥2 falls	Cross-sectional	≥ 2 falls (n=10) (6:4) 71.4±2.9	NA	KE: 2.0±0.3 Nm/kg	RFD200KE: 3.5±2.0 Nm/s/kg	≥ 2 falls within the last year 30 s Chair stand: 19.6±7.6 Usual gait speed: 1.37±0.36 m/s One leg standing test: 80.1±51.6 s
		1 fall (n=24) (11:13) 71.2±3.7	NA	KE: 2.1±0.7 Nm/kg	RFD200KE: 6.5±3.6 Nm/s/kg*	1 fall within the last year 30 s chair stand: 22.6±6.6 Usual gait speed:1.31±0.18 m/s One leg standing test: 93.0±41.3 s
		Non-fallers (n=88) (43:45) 71.3±4.7	NA	KE: 2.1±0.6 Nm/kg	RFD200KE: 5.8±2.7 Nm/s/kg*	No history of falls 30 s chair stand: 22.1±7.5 Usual gait speed:1.37±0.19 m/s One leg standing test: 82.4±41.8 s
Data presented as mean±SD	*Different from multiple falls groups		Equipment: Mobie hand-held dynamometer
LaRoche et al. 2010 Elderly fallers vs. non-fallers	Cross-sectional	Fallers (n=11) (11:0) 71.3±5.4	NA	KE: 1.49±0.46 Nm/kg KF: 0.59±0.73 Nm/kg DF: 0.32±0.06 Nm/kg PF: 0.76±0.14 Nm/kg *(Combined measure)	RFD200KE: 6.97±2.9 Nm/s/kg   KF: 4.02±2.17 Nm/s/kg   DF: 1.57±0.36 Nm/s/kg   PF: 3.18±1.14 Nm/s/kg	≥3 falls within the last year
		Non-fallers (n=12) (12:0) 71.2±6.2	NA	KE: 1.72±0.56 Nm/kg KF: 0.72±0.25 Nm/kg DF: 0.38±0.09 Nm/kg PF: 1.04±0.27 Nm/kg	RFD200KE: 6.90±3.86 Nm/s/kg   KF: 4.50±2.67 Nm/s/kg   DF: 1.93±0.55 Nm/s/kg   PF: 4.12±1.89 Nm/s/kg	No history of unexplained falls
Data presented as mean±SD	*Difference in composite Z-score between groups		Equipment: Humac Norm dynamometer
LaRoche et al. 2011 Low vs. normal strength in elderly	Cross-sectional	Low strength (n=13) (13:0) 71.2±4.3	NA	KE: 1.16±0.16 Nm/kg* KF: 0.69±0.18 Nm/kg PF: 0.55±0.25 Nm/kg* DF: 0.34±0.06 Nm/kg	RFD200KE: 6.25±2.62 Nm/kg/s*  KF: 3.32±1.44 Nm/kg/s*  PF: 2.84±1.88 Nm/kg/s  DF: 1.43±0.59 Nm/kg/s	KE torque <1.5 Nm/kg SPPB: 10.7±1.3 Balance score: 3.8±0.6 Habitual gait speed: 1.12±0.19 m/s* Chair rise: 11.5±1.15 s
		Normal strength (n=11) (11:0) 72.1±5.2	NA	KE: 1.65±0.13 Nm/kg KF: 0.83±0.14 Nm/kg PF: 0.85±0.24 Nm/kg DF: 0.39±0.11 Nm/kg	RFD200KE: 9.51±3.29 Nm/kg/s   KF: 4.77±1.19 Nm/kg/s   PF: 3.64±1.41 Nm/kg/s   DF: 1.92±0.59 Nm/kg/s	KE torque >1.5 Nm/kg SPPB: 11.4±0.7 Balance score: 3.9±0.3 Habitual gait speed: 1.32±0.25 m/s Chair rise: 10.3±1.4 s
Data presented as mean±SD	*Different from Normal strength		Equipment: Humac Norm dynamometer
Lopez et al. 2017 Elderly	Cross-sectional	Elderly sedentary men (n=50) (0:50) 66±5.4	NA	NR	RFD50KE: 1.24±0.62 Nm/s RFD100KE: 1.01±0.40 Nm/s	30s STS: 16.7±2.6
Data presented as mean±SD	Equipment: Cybex Norm dynamometer
Mackey et al. 2006 Elderly vs. young	Cross-sectional	Elderly (n=25) (25:0) 78±7	NA	VGRFPF, static: 0.89±0.12 Nm/kg×m  PF, dynamic: 0.96±0.16 Nm/kg×m*	VGRFD(0-85%MVC)PF, dynamic: 5.57±1.98 Nm/s×kg×m	Balance recovery trials:  Static recovery angle: 13.1±2.4°*  Dynamic recovery angle: 4.6±1.8°*
		Young (n=25) (25:0) 25±4	NA	VGRFPF, static: 0.93±0.06 Nm/kg×m   PF, dynamic: 1.04±0.15 Nm/kg×m NB! Strength data is measured during balance recovery trials.	VGRFD(0-85%MVC)PF Dynamic: 6.60±2.49 Nm/s×kg×m	Balance recovery trials:  Static recovery angle: 16.3±1.5°  Dynamic recovery angle: 7.2±1.2°
Data presented as mean±SD	*Different from Young	Equipment: Bertec force plate
Morcelli et al. 2016a+b Elderly fallers vs. non-fallers vs. healthy young females	Cross-sectional	Fallers (n=20) (20:0) 68.9±6.5	NA	HE: 1.23±0.46 Nm/kg HF: 0.69±0.17 Nm/kg HAb: 0.72±0.22 Nm/kg HAd: 0.52±0.19 Nm/kg	RFD50HE: 1.68±0.82 Nm/s/kg   HF: 1.71±0.90 Nm/s/kg   HAb: 1.52±0.75 Nm/s/kg   HAd: 0.90±0.59 Nm/s/kg RFD100HE: 1.91±0.95 Nm/s/kg   HF: 1.75±0.88 Nm/s/kg   HAb: 1.74±0.87 Nm/s/kg   HAd: 0.99±0.62 Nm/s/kg RFD150HE: 1.89±0.92 Nm/s/kg   HF: 1.43±0.72 Nm/s/kg   HAb: 1.64±0.80 Nm/s/kg   HAd: 0.86±0.55 Nm/s/kg RFD200HE: 1.23±0.46 Nm/s/kg   HF: 1.06±0.44 Nm/s/kg   HAb: 1.27±0.67 Nm/s/kg   HAd: 0.68±0.36 Nm/s/kg	Fall(s) in the year before evaluation
		Non-fallers (n=24) (24:0) 65.5±6.16	NA	HE: 1.41±0.35 Nm/kg HF: 0.76±0.1 Nm/kg HAb: 0.82±0.15 Nm/kg HAd: 0.57±0.14 Nm/kg	RFD50HE: 2.36±0.96 Nm/s/kg   HF: 1.80±0.56 Nm/s/kg   HAb: 1.85±0.70 Nm/s/kg   HAd: 1.22±0.61 Nm/s/kg RFD100HE: 2.58±1.01Nm/s/kg   HF: 1.97±0.61 Nm/s/kg   HAb: 2.32±0.80 Nm/s/kg   HAd: 1.37±0.67 Nm/s/kg RFD150HE: 2.38±0.87 Nm/s/kg   HF: 1.77±0.52 Nm/s/kg   HAb: 2.37±0.74 Nm/s/kg§   HAd: 1.22±0.53 Nm/s/kg RFD200HE: 1.95±0.63 Nm/s/kg   HF: 1.36±0.41 Nm/s/kg   HAb: 1.94±0.66 Nm/s/kg§ HAd: 0.89±0.35 Nm/s/kg	No falls in the year before evaluation
		Young (n=18) (18:0) 21.8±2.1	NA	HE: 2.61±0.58 Nm/kg*§ HF: 1.24±0.15 Nm/kg*§ HAb: 1.37±0.26 Nm/kg*§ HAd: 1.14±0.32 Nm/kg*§	RFD50HE: 6.53±2.13 Nm/s/kg*§   HF: 4.70±1.53 Nm/s/kg*§   HAb: 5.64±1.82 Nm/s/kg*§   HAd: 3.99±1.78 Nm/s/kg*§ RFD100 HE: 7.16±2.38 Nm/s/kg*§   HF: 4.93±1.36 Nm/s/kg*§   HAb: 6.32±1.80 Nm/s/kg*§   HAd: 4.33±1.78 Nm/s/kg*§ RFD150HE: 6.33±2.07 Nm/s/kg*§   HF: 3.83±0.85 Nm/s/kg*§   HAb: 5.07±1.27 Nm/s/kg*§   HAd: 3.55±1.29 Nm/s/kg*§ RFD200HE: 4.75±1.53 Nm/s/kg*§   HF: 2.25±0.87 Nm/s/kg*§   HAb: 2.30±1.20 Nm/s/kg§   HAd: 2.01±0.77 Nm/s/kg*§	__
Data presented as mean±SD	*Difference compared to Elderly non-fallers §Difference compared to Elderly fallers		Equipment: Biodex dynamometer
Palmer et al. 2015 Elderly fallers vs. non-fallers	Cross-sectional	Fallers (n=6) (6:0) 72.67±6.89	NA	HE: 7.96±3.04 Nm	RFD50HE: 37.43±23.95 Nm/s RFD100-200HE: 28.73±17.70 Nm/s	>1 fall in the last 12 mo
		Non-fallers (n=9) (9:0) 71.44±6.95	NA	HE: 11.16±4.59 Nm	RFD50HE: 80.86±48.12 Nm/s* RFD100-200HE: 34.28±18.56 Nm/s	No falls in the last 12 mo
Data presented as mean±SD	*Difference compared to fallers		Equipment: Load cell
Palmer et al. 2016 Elderly higher vs. lower functioning	Cross sectional	Higher functioning (n=9) (6:3) 87.1±6.0	NA	KE: 51.2±26.0 Nm KF: 35.1±13.2 Nm	RFD50KE: 241.5±111.5 Nm/s  KF: 238.4±117.1 Nm/s * (collapsed across muscle) RFD200KE: 121.3±49.1 Nm/s   KF: 109.3±47.9 Nm/s	Able to successfully rise from chair
		Lower functioning (n=6) (3:3) 89.2±6.0	NA	KE: 39.3±16.3 Nm KF: 32.4±15.1 Nm	RFD50KE: 110.9±56.8 Nm/s   KF: 129.8±62.1 Nm/s RFD200KE: 89.5±56.9 Nm/s   KF: 101.5±68.0 Nm/s	Unable to successfully rise from chair
Data presented as mean±SD	*Different compared with Lower functioning when collapsed across muscle			Equipment: Load cell
Palmer et al. 2017 Young vs. old	Cross-sectional	Elderly (n=11) (11:0) 67±8	NA	HE: 70.28±31.02 Nm*	RFD50HE: 254.30±105.87 Nm/s* RFD200HE: 180.16±69.0 Nm/s*	OSI: 0.69±0.19*
		Young (n=11) (11:0) 26±8	NA	HE: 94.89±23.95 Nm	RFD50HE: 364.14±121.74 Nm/s RFD200HE: 274.29±82.17 Nm/s	OSI: 0.48±0.16
Data presented as mean±SD	*Different from Young females		Equipment: Load cell
Pijnappels et al. 2008 Elderly fallers vs. non-fallers	Cross-sectional	Fallers (n=7) (7:0) 71±4.5	NA	KE: 1.44±0.41 Nm/kg* PF: 1.41±0.39 Nm/kg* Leg press: 11.74±1.5 N/kg*	RFD100KE: 5.02±1.71 Nm/s/kg*   PF: 3.81±1.29 Nm/s/kg*   Leg press: 31.16±17.52 N/s/kg	Fully supported by safety harness in more than half of the tripping trials.
		Non-fallers (n=10) (3:7) 71±4.5	NA	KE: 2.07±0.45 Nm/kg PF: 1.83±0.36 Nm/kg Leg press: 15.61±2.06 N/kg	RFD100KE: 8.62±3.12 Nm/s/kg   PF: 6.01±2.1 Nm/s/kg   Leg press: 51.47±28.53 N/s/kg	Never fully supported by the safety harness.
Data presented as mean±SD Strength data extracted from figure		*Different from Non-fallers	Equipment: Cybex Norm dynamometer
Rech et al. 2014 Elderly	Cross-sectional	Active old women (n=45) (45:0) 70.28±6.2	NA	KE: 108.09±28.7 Nm	RFD50KE: 0.47±0.25 Nm/ms RFD100KE: 0.45±0.24 Nm/ms RFD250KE: 0.29±0.11 Nm/ms RFD300KE: 0.25±0.09 Nm/ms	30s STS: 12.9±2.3 Usual gait speed: 1.3±0.2 m/s
Data presented as mean±SD	Equipment: Cybex Norm dynamometer
Seynnes et al. 2005 Elderly	Cross-sectional	Elderly (n=19) (19:0) 77.9±1.2	NA	KE: 108.1±5.1 Nm	RFDmaxKE: 126.0±18.7 Nm/s	Chair-rise: 0.64±0.02 s Stairclimbing power: 224.71±9.78 W 6MWT: 348.6±8.2 m
Data presented as mean±SD	Equipment: Biodex dynamometer
Sundstrup et al. 2010 Elderly trained vs. untrained vs. young untrained	Cross-sectional	Football-trained elderly (n=10) (0:10) 69.6±1.4	NA	KE: 2.32±0.12 Nm/kg	RFD30KE: 10.27±0.77 Nm/s/kg* RFD100KE: 13.34±0.68 Nm/s/kg* RFD200KE: 9.26±0.54 Nm/s/kg*	Flamingo balance test: 15.5±1.0*
		Untrained elderly (n=8) (0:8) 70.5±1.0	NA	KE: 2.21±0.18 Nm/kg	RFD30KE: 6.88±1.15 Nm/s/kg RFD100KE: 8.08±1.30 Nm/s/kg RFD200KE: 6.68±0.87 Nm/s/kg	Flamingo balance test: 33.2±1.9
		Untrained young (n=49) (0:49) 32.4±0.9	NA	KE: 2.90±0.07 Nm/kg*§	RFD30KE: 11.32±0.43 Nm/s/kg* RFD100KE: 14.29±0.43 Nm/s/kg* RFD200KE: 10.11±0.32 Nm/s/kg*	Flamingo balance test: 15.0±0.6*
Data presented as mean±SE	*Different from Untrained elderly . §Different from Football-trained elderly			Equipment: KinCom dynamometer
Thompson et al. 2018 Young vs. old	Cross-sectional	Elderly (n=18) (10:8) 71.1±5.9	NA	Squat: 432.7±47.46 Nm* KE+KF: 150.8±62.98 Nm*	RFD50Squat: 590.2±193.35 Nm/s*   KE+KF: 225.8±188.95 Nm/s RFD200Squat: 1073.5±653.4 Nm/s*   KE+KF: 529.2±261 Nm/s*	NR 10 m walk 400 m walk Chair stand
		Young (n=20) (10:10) 21.9±2.6	NA	Squat: 635.3±169.78 Nm KE+KF: 209.4±151.52 Nm	RFD50Squat: 1076.6±681.11 Nm/s   KE+KF: 275.3±355.94 Nm/s RFD200Squat: 1872.1±674.8 Nm/s   KE+KF: 728.9±243.4 Nm/s	NR 10 m walk 400 m walk Chair stand
Data presented as mean±SD	*Different from young	Equipment: Biodex dynamometer
Unhjem et al. 2019 Young vs. old	Cross-sectional	Sedentary elderly (n=10) (0:10) 71±4	NA	Leg press, dyn: 106±3 kg*	RFD30Leg pres, dyn: 1610 N/s RFD50Leg pres, dyn: 1946 N/s RFD100Leg pres, dyn: 2389 N/s RFD150Leg pres, dyn: 2554 N/s RFD200Leg pres, dyn: 2495 N/s	Habitual walking speed: 1.26 m/s* Chair rise: 9.7 s* Stairclimbing: 498 W* One leg-standing:  Without multitasking: 7.07 cm/s*  With multitasking: 9.43 cm/s*
		Active elderly (n=11) (0:11) 73±6	NA	Leg press, dyn: 128±4 kg*§	RFD30Leg pres, dyn: 2035 N/s RFD50Leg pres, dyn: 2690 N/s RFD100Leg pres, dyn: 3433 N/s RFD150Leg pres, dyn: 3478 N/s RFD200Leg pres, dyn: 3097 N/s	Habitual walking speed: 1.56 m/s§ Chair rise: 8.6 s* Stairclimbing: 554 W* One leg-standing:  Without multitasking: 5.37 cm/s*  With multitasking: 7.21 cm/s*
		Old master athletes (n=11) (0:11) 71±4	NA	Leg press, dyn: 186±10 kg*#§	RFD30Leg pres, dyn: 4389 N/s RFD50Leg pres, dyn: 5451 N/s RFD100Leg pres, dyn: 5999 N/s RFD150Leg pres, dyn: 5636 N/s RFD200Leg pres, dyn: 4961 N/s	Habitual walking speed: 1.49 m/s§ Chair rise: 6.2 s#§ Stairclimbing: 701 W*#§ One leg-standing:  Without multitasking: 6.46 cm/s*  With multitasking: 8.12 cm/s*
		Young (n=9) (0:9) 22±2	NA	Leg press, dyn: 147±7 kg	RFD30Leg pres, dyn: 3292 N/s RFD50Leg pres, dyn: 4053 N/s RFD100Leg pres, dyn: 5008 N/s RFD150Leg pres, dyn: 4961 N/s RFD200Leg pres, dyn: 4491 N/s	Habitual walking speed: 1.62 m/s Chair rise: 6.5 s Stairclimbing: 879W One leg-standing:  Without multitasking: 4.02 cm/S  With multitasking: 4.45 cm/s
Data presented as mean±SD	*Different from young. #Different from active older adults. §Different from sedentary older adults				Equipment: Force plate
Parkinson Disease
Malling et al. 2016 PD patients vs. healthy controls	Intervention (controlled) 8 w control period followed by motor ‘reactive’ training + aerobic training (8 w, 24 sess.)	PD patients (n=13) (0:13) 63	Hoehn & Yahr staging: 2.1	NR	VGRFD(30-70%MVC)STS, up:  pre: 10.66±0.93 ×BW/s  post4w: 11.33±0.96 ×BW/s  post8w: 13.67±2.25×BW/s  ∆: 3.01 ×BW/s (28.24%) VGRFD(30-70%MVC)STS, down:  pre: 7.94±0.8 ×BW/s  post4w: 8.85±0.75 ×BW/s*  post8w: 9.7±0.88 ×BW/s*  ∆: 1.76 ×BW/s (23%) VGRFD(0-70%MVC)DPB, lat:  precontrol period: 2.06±0.3 ×BW/s  pre: 2.02±0.37 ×BW/s  post8w: 2.36±0.36 ×BW/s*  ∆8w: 0.34 ×BW/s (16%) VGRFD(0-70%MVC)DPB, med:  precontrol period: 1.85±0.35 ×BW/s  pre: 1.71±0.33 ×BW/s  post8w: 2.04±0.36 ×BW/s*  ∆8w: 0.33 ×BW/s (19%)	5STS:  Completion time:   pre: 9.93±1.08 s   post4w: 8.14±0.82 s*   post8w: 7.51±0.88 s*   ∆8w: -2.42 s (-24%)  Standing-time:   pre: 0.73±0.11 s   post4w: 0.58±0.06 s*   post8w: 0.54±0.05 s*   ∆8w: -0.19 s (-27%)  Sitting-time:   pre: 0.51±0.06 s   post4w: 0.37±0.05 s*   post8w: 0.32±0.05 s*   ∆8w: -0.19 s (-36%) Dynamic postural balance:  Completion time:   precontrol period: 24.35±4.03 s   pre: 20.94±3.41 s   post8w: 19.32±3.26 s  ∆8w: -1.62 s (-7.73%)  Flatness:   precontrol period: 25.23±5.43 N   pre: 23.06±4.01 N   post8w: 23.14±3.51 N   ∆8w: 0.08 s (0.35%)
		Healthy controls (n=17) (0:17) 58	NA	NR	VGRFD(30-70%MVC)STS, up:  pre: 14.83±0.41×BW/s * VGRFD(30-70%MVC)STS, down:  pre: 10.26±0.35×BW/s * VGRFD(0-70%MVC)DPB, lat:  precontrol period: 3.78±0.3 ×BW/s§ VGRFD(0-70%MVC)DPB, med:  precontrol period: 3.28±0.27×BW/s§	5STS: Completion time: pre: 7.33±0.36 s* Standing-time: pre: 0.55±0.02 s* Sitting-time: pre: 0.39±0.03 s Dynamic postural balance:  Completion time:   precontrol period: 12.32±1.13 s§  Flatness:   rowspan="2"precontrol period: 40.60±5.35 N§
Data presented as mean±SE RFD and functional data extracted from figure		*Different compared to pre-values after control period for PD patients §Different compared to pre-values before control period for PD patients			Equipment: AMTI force plate
Noorvee et al. 2006 PD patients vs. healthy controls	Cross-sectional	PD patients (n=12) (5:7) 67.4±1.2	Hoehn & Yahr staging: II-III	KE: 292.77±25.3 N	RFD200KE: 832.84±109.97 N/s*	Postural sway test:  Eyes open: 7.25±1.76  Eyes closed: 6.68±1.2 6 m walk: 0.99±0.06 m/s* UPDRS motor: NR
		Healthy controls (n=12) (5:7) 66.8±1.1	NA	KE: 317.17±23.49 N	RFD200KE: 1129.03±105.57 N/s	Postural sway test:  Eyes open: 5.92±0.37  Eyes closed: 5.15±10.66 6 m walk: 1.14±0.03 m/s UPDRS motor: NR
Data presented as mean±SE Strength data extracted from figure		*Different from Healthy controls		Equipment: Custom-made dynamometric chair
Pääsuke et al. 2002 PD patients vs. healthy controls	Cross-sectional	PD patients (n=14) (14:0) 72.6±2.2	10.3±1.2 y Hoehn & Yahr staging: I-III	KE  Absolute   Right: 224.32±14.87 N*   Left: 208.11±13.51 N*  Normalized (relative to BW)   Right: 39.22±2.29%   Left: 35.15±2.3% VGRFSTS:  Right: 371.74±20.83 N  Left: 344.98±12.64 N	RFDpeakKE:  Right: 867.83±103.5 N/s*  Left: 835.99±79.62 N/s* VGRFDmaxSTS:  Right: 874.25±125.75 N/s*  Left: 838.32±179.64 N/s*	Chair rise: 3.42±0.14 s*
		Healthy controls (n=12) (12:0) 72.8±0.8	NA	KE  Absolute   Right: 267.57±18.92 N   Left: 263.51±13.52 N  Normalized (relative to BW)   Right: 42.40±2.65%   Left: 41.69±1.94% VGRFSTS:  Right: 364.31±29.74 N  Left: 358.36±22.31 N	RFDpeakKE:  Right: 1504.78±175.16 N/s  Left: 1449.04±167.2 N/s VGRFDmaxSTS:  Right: 1305.39±215.57 N/s  Left: 1239.52±227.55 N/s	Chair rise: 2.29±0.14 s
Data presented as mean±SE Strength data extracted from figure		*Different from Healthy controls		Equipment: Dynamometer & VISTI force plates
Pääsuke et al. 2004 PD patients vs. healthy controls	Cross-sectional	PD patients (n=12) (12:0) 74.3±6.9	10.7±4.5 y Hoehn & Yahr staging: I-III	Bilateral KE  Absolute: 711.94±290.4 N*  Normalized: 11.35±4.96 N/kg* Unilateral KE  More affected: 483.87±169.35 N  Less affected: 548.38±169.36 N VGRFSTS: 705.1±91.43 N	RFDpeakKE bilateral: 2625±1521.15 N/s* VGRFDmaxSTS: 2391.13±818.55 N/s*	Chair rise: 2.45±0.24 s*
		Healthy controls (n=16) (16:0) 71.7±4.4	NA	Bilateral KE  Absolut: 1147.54±515.22 N  Normalized: 17.45±5.55 N/kg Unilateral KE  Dominant: 802.42±330.64 N  Non-dominant: 733.87±358.9 N VGRFSTS: 735.03±92.3 N	RFDpeakKE bilateral: 4092.31±1709.61 N/s VGRFDmaxSTS: 3411.29±975.81 N/s	Chair rise: 1.86±0.3 s
Data presented as mean±SD Strength data extracted from figure		*Different from Healthy controls		Equipment: Dynamometer & VISTI force plates
Multiple Sclerosis
Kjølhede et al. 2015a MS patients	Cross-sectional	MS patients (n=25) (26:8) 43.3±8.2	EDSS: 2.9	Stronger leg:  KE: 2.30±0.53 Nm/kg  KF: 0.97±0.29 Nm/kg Weaker leg:  KE: 1.99±0.49 Nm/kg*  KF: 0.81±0.29 Nm/kg*	Stronger leg:  RFDmaxKE: 13.43±3.46 Nm/kg/s  RFDmaxKF: 8.42±1.35 Nm/kg/s  RFD200KE: 7.35±2.49 Nm/kg/s  RFD200KF: 2.98±0.99 Nm/kg/s Weaker leg:  RFDmaxKE: 12.65±3.32 Nm/kg/s*  RFDmaxKF: 8.09±1.38 Nm/kg/s*  RFD200KE: 6.56±2.20 Nm/kg/s*  RFD200KF: 2.39±0.92 Nm/kg/s*	T25FWT: 1.72±0.31 m/s 2MWT: 1.64±0.33 m/s 5STS: 9.48±2.70 s Stairclimbing: 10.64 s
Data presented as mean±SD	*Different from stronger leg		Equipment: Humac Norm dynamometer
Stroke
Nadeau et al. 1997 Stroke patients	Cross-sectional	Hemiparetic stroke patients (n=16) (4:12) 47.9±15.6	43.9±36.5 mo Fugl-Meyer Assessment: 184.3	PF: 50.8±24.4 Nm	RFDpeakPF: 110.5±56.7 Nm/s	TUG: 9.20±2.40 s Comfortable walking speed: 0.76±0.27 m/s Maximal safe walking speed: 1.08±0.33 m/s
Data presented as mean±SD	Equipment: Biodex dynamometer
Pohl et al. 2002 Stroke patients	Cross-sectional	Stroke patients (n=83) (39:44) 70.3±9.8	78.6±27.4 d Fugl-Meyer Assessment: 23.7±3.7	Affected knee:  KE: 53.55±24.27 Nm Less affected knee:  KE: 52.58±25.62 Nm	Affected knee:  RFD150KE: 66.02±69.42 Nm/s Less affected knee:  RFD150KE: 81.76±77.01 Nm/s	10 m walking speed: 63.2±25.9 cm/s
Data presented as mean±SD	Equipment: Cybex dynamometer
Takeda et al. 2018 Stroke patients	Cross-sectional	Chronic stroke patients (n=20) (3:17) 63.6±10.1	5.7 y Fugl-Meyer Assessment: 17.9±6.5	NR	Affected limb:  RFD50KE: 6.48±4.62 N/kg/s*  RFD100KE: 8.82±5.75 N/kg/s*  RFD200KE: 6.93±3.99 N/kg/s*  RFD300KE: 6.08±3.39 N/kg/s* Non-affected limb:  RFD50KE: 14.96±13.81 N/kg/s  RFD100KE: 15.23±9.10 N/kg/s  RFD200KE: 13.14±7.74 N/kg/s  RFD300KE: 11.70±5.97 N/kg/s	Walking speed:  Maximum: 0.75 m/s  Comfortable: 0.58 m/s
Data presented as mean±SD	*Different from non-affected limb		Equipment: Handheld dynamometer

NA: not applicable; NR: not reported; SD: standard deviation; IQR: interquartile range; MS: Multiple Sclerosis; PD: Parkinson Disease; EDSS: Expanded Disability Status Scale; UPDRS: Unified Parkinson’s Disease Rating Scale; RCT: Randomized controlled trial; QF: Quadriceps Femoris; KE: Knee extension; KF: Knee flexion; DF: Ankle dorsiflexion; PF: Plantarflexion; HE: Hip extension; HF: Hip flexion; Ham: Hamstrings; HAd: Hip adductor; HAb: Hip abductor; LE: leg extension; Con: concentric; Ecc: eccentric; Dyn: dynamic; BW: body weight; RFD: rate of moment development; VGRF: Vertical ground reaction force; VGRFD: Vertical ground reaction force development; T25FWT: Timed 25 Foot Walk Test; 2MWT: Two-minute Walk Test; 6MWT: Six-minute Walk Test; 30s STS: Sit-to-stand movement; BBT: Berg Balance Test; BBS: Berg Balance Scale, OSI: overall stability index; TUG: Time up-and-go; SPPB: Short Physical Performance Battery; COP: center of pressure; LCOP: total length of the COP trajectory; ACOP: area of the rectangle circumscribing the COP trajectory; ETGUG: expandend timed get up-and-go; FAB scale: Fullerton Advanced Balance scale; FAC: Functional Ambulation Category; AP: anteroposterior direction; ML: mediolateral direction; CTSIB: Clinical test of Sensory Interaction on Balance Test; DPB: Dynamic postural balance test. Unless else is noted, all strength data was measured during isometric contractions. Italic values indicates that the value was calculated by the investigator. Red text denote excluded outcomes or studies from the main analysis.

In addition to the qualitative analysis (summary of identified studies and their data), we also performed quantitative analysis by calculating sample size weighted averages across selected studies. These data are presented as mean±CI95%. Also, within- and between-outcome analyses were carried out by using linear mixed model, with study set as random effect and outcome (Fmax, RFD, Function) as fixed effect. Simple unadjusted regression analyses were carried out to examine associations between parallel observations of deficits (Part 1). All statistical analyses were carried out using STATA (STATA/IC 14.2, StataCorp, College Station, Texas, USA), while graphical illustrations were created using GraphPad Prism 7.0 (GraphPad Software, La Jolla, California, USA, www.graphpad.com).

Results

General study characteristics

The literature search yielded N=1399 potential articles of which N=32 articles (n=1087 study participants) were ultimately included (after removal of duplicates, screening of title and abstract, full text reading, check of reference lists) (Figure 1). Across the different populations, N=24 articles (n=830 subjects) evaluated aging[48-71], N=4 articles (n=108 subjects) evaluated PD[30,44,72,73], N=1 article (n=34 subjects) evaluated MS[33], and N=3 articles (n=115 subjects) evaluated Stroke[45,74,75]. No articles investigating AD met the inclusion criteria. Except for one study[56], large-scale studies having >100 participants were lacking as sample sizes ranged from 15 to 83 subjects. All studies reported data as mean and standard deviations or standard errors. Fmax was presented as Nm, Nm/kg, Nm/m/kg, N, N/kg, or kg, and RFD values as Nm/s, Nm/s/kg, N/s, N/s/kg, Nm/s/kg/m, N/ms, Nm/ms, or ×BW/s. Regarding the analysis of deficits (Part 1), all strength data was normalised to body weight, except for strength outcomes measured as vertical ground reaction force or vertical ground reaction RFD. Supplementary Table 2 contain detailed information of the identified studies including all extracted outcomes of Fmax, RFD, and physical function.

Figure 1.

Study flowchart.

Large heterogeneity was identified in the assessment of physical function and lower extremity Fmax and RFD (in particular) of the included studies. Specifically, a wide range of muscle actions was used to generate lower extremity Fmax and RFD, but most often involving knee extension/flexion, hip flexion/extension/abduction/adduction, and ankle plantar/dorsi flexion (Supplementary Table 2, Table 1, Table 2). While RFD was derived from isometric Fmax (also termed maximal voluntary isometric contraction, MVC) in most studies, some also derived it from dynamic Fmax during plantar flexion[50], leg press[71], balance trials[59,73], sit-to-stand movements[30,72,73], lateral stepping[54], squat jump , and isokinetic (60°/s), concentric (90°/s and 180°/s), and eccentric contractions[51]. Measures of RFD were quite divergent, but most often presented as RFD_50ms, RFD_100ms, RFD_200ms, and RFD_max (Supplementary Table 2). Also, a wide range of different tests of physical function was applied, with data being summarized according to the classification outlined in Methods (i.e. comprising short and long walking tests, chair rise test tests, stair climbing, TUG, SPPB, retrospective fall history) (Supplementary Table 2). As for the different balance tests, these were deemed too divergent that it left us unable to summarize data. Consequently, balance data was excluded from further data analysis (for transparency, specific information on these tests along with data are reported in Supplementary Table 2).

Table 2.

Associations (Part 2).

	Study		Associations (r2-values)		Muscle actions	Assessment details
			Fmax-Function	RFD-Function
Older adults	Altubasi et al. 2015		n=21	0.05	0.12	Knee ext	RFDmax Short walkusual+max, Stair climb, TUG
	Bento et al. 2010		n=31		0.02	Knee flex	RFD20-80% Number of falls
	Crocket et al. 2015		n=29		0.08	Knee ext	RFDmax Chair rise
	Hester et al. 2019		n=26		0.25	Plantar flex (dyn)	VGRFD Chair rise, Short walkusual+max, TUG
	LaRoche et al. 2011		n=24	0.25	0.22	Dorsi+Plantar flex Knee ext+flex	RFD200ms Long walkusual+max
	Lopez et al. 2017		n=50		0.16	Knee ext	RFD50ms, RFD100ms Chair rise
	Rech et al. 2014		n=45	0.15	0.11	Knee ext	RFD50ms, RFD100ms, RFD250ms, RFD300ms Chair rise, Short walkusual
	Seynnes et al. 2005		n=19	0.20	0.05	Knee ext	RFDmax Chair rise, Long walkmax, Stair climb
	Thompson et al. 2018		n=18	0.29	0.19	Knee ext+flex, Squat	RFD50ms, RFD200ms Chair rise, Long walkmax, Short walkmax
	Unhjem et al. 2019		n=41	0.30	0.27	Leg press (dyn)	RFD30ms, RFD50ms, RFD100ms, RFD150ms, RFD200ms Chair rise
	Weighted mean 95% CI [lower : upper]	Participants Studies	n=295 N=10	0.19 [0.12 : 0.27]	0.14 [0.09 : 0.19]
PD patients	Pääsuke et al. 2002		n=14	0.19	0.09	Knee ext	RFDmax Chair rise
	Pääsuke et al. 2004		n=12	0.22	0.09	Knee ext (bilat.)	RFDmax Number of falls
	Weighted mean 95% CI [lower : upper]	Participants Studies	n=26 N=2	0.20 [0.20 : 0.21]	0.09 [0.08 : 0.10]
MS patients	Kjølhede et al. 2015a		n=34	0.13	0.11	Knee ext+flex	RFD200ms, RFDmax Chair rise, Long walkmax, Short walkmax, Stair climb
Stroke patients	Nadeau et al. 1997		n=12	0.06	0.09	Plantar flex	RFDpeak Short walkusual+max, TUG
	Pohl et al. 2002		n=83	0.04	0.09	Knee ext	RFD150ms Short walkusual
	Takeda et al. 2018		n=20	0.20	0.24	Knee ext	RFD50ms, RFD100ms, RFD200ms, RFD300ms Short walkusual+max
	Weighted mean 95% CI [lower : upper]	Participants Studies	n=115 N=3	0.07 [0.03 : 0.11]	0.11 [0.04 : 0.19]
All	Weighted mean 95% CI [lower : upper]	Participants Studies	n=470 N=16	0.15 [0.10 : 0.20]	0.13 [0.09 : 0.17]

TUG: timed up and go. VGRFD: vertical ground reaction RFD (derived from force plate). In studies reporting data on more than one association between Fmax or RFD (based on a mean RFD value in studies reporting data on more than one RFD measure) and physical function, mean association values of Fmax-Function and RFD-Function, respectively, were calculated.

Part 1 – Parallel observations of deficits in Fmax, RFD, and physical function

Data extraction identified studies reporting parallel observations of Fmax and RFD deficits in older fallers vs. non-fallers, older lower functioning vs. higher functioning, older vs. young, and PD patients vs. healthy controls. Deficits were observed in Fmax (ranging from -11% to -27%, p<0.001), RFD (ranging from -18% to -41%, p<0.001) and Function (ranging from -13% to -31%, p<0.001) in each separate study population (specific results are displayed in Table 1 and Figure 2A).

Figure 2A.

Parallel observations of deficits in Fmax, RFD, and physical function (Part 1). Sample size weighted average deficits in Fmax, RFD, and physical function across all identified studies. Data are presented as mean±CI95%. See Table 1 for further data details.

Across all study populations (comprising N=20 studies/n=706 participants in Fmax, N=22/n=817 in RFD, and N=7/n=193 in Function), deficits were greater in RFD vs. Fmax (-31% [-41:-20] vs. -21% [-29:-12], respectively; difference=10.9% point [5.5:16.3], p<0.000) (Table 1, Figure 2A) and greater in Function vs. Fmax (-26% [-36:-16] vs. -21% [-29:-12], respectively; difference=14.3% point [5.5:23.1], p=0.002), but comparable between RFD and Function (-31% [-41:-20] vs. -26% [-36:-16], respectively; difference=3.3% point [-5.1:11.8], p=0.441) (Table 1, Figure 2A). Across these studies, no calculated associations were observed between deficits in Fmax and Function (r²=0.09 [-0.63:0.87], p=0.630, N=5) or between deficits in RFD and Function (r²=0.34 [-0.09:0.86], p=0.167, N=7). A calculated association was, however, observed between deficits in Fmax and RFD (r²=0.81 [0.57:0.92], p<0.000, N=20).

Deficits in Fmax and RFD across separate muscle groups and across separate contraction modes (isometric and dynamic; early and late RFD) are shown in Figure 2B. Deficits were greater in RFD (N=4) vs. Fmax (N=5) during knee+hip extensor muscle actions (i.e. comprising leg press, squat and chair rise) (difference=21.8% point [-27.6:-16.1], p<0.000), but not in any other of these outcomes (Figure 2B left). Also, deficits were greater in RFD (N=19) vs. Fmax (N=18) during isometric muscle contractions (difference=7.5% point [-11.8:-3.2], p=0.001), and greater in RFD (N=4) vs. Fmax (N=6) during dynamic muscle contractions (difference=22.9% point [-25.2:-20.5], p<0.000) (Figure 2B middle). However, derived from isometric muscle contractions (if reported), deficits did not differ between early RFD (N=10) and late RFD (N=11) (-38% [-58:-18] vs. -28% [-43:-13], respectively; difference=2.8% point [-3.5:9.0], p=0.385) (Figure 2B right).

Figure 2B.

Parallel observations of deficits in Fmax and RFD across muscle groups and contraction mode. Sample size weighted average deficits in Fmax and RFD across muscle groups as well as contraction mode (including early and late RFD derived from isometric contractions) from all identified studies. “Hip + Knee Ext” is comprised of leg press, squat, and chair rise. The presented data support and elaborate Figure 2A and Table 1. Data are presented as mean±CI95%.

Part 2 – Associations between RFD or Fmax and physical function

Data extraction identified studies reporting associations between Fmax or RFD and physical function in older adults as well as in PD, MS, and Stroke patients. Reported associations were observed for Fmax-Function (ranging from r²=0.07 to r²=0.20) and RFD-Function (ranging from r²=0.09 to r²=0.14) in each separate study population (specific results are displayed in Table 2 and Figure 2C).

Figure 2C.

Associations between RFD or Fmax and Physical function (Part 2). Sample size weighted average associations between Fmax or RFD and physical function (denoted Fmax-Function and RFD-Function, respectively) being reported across all selected studies. Data are presented as mean±CI95%. See Table 2 for further data details.

Across all studies (comprising N=12/n=334 in Fmax-Function and N=16/n=470 in RFD-Function), reported associations between Fmax-Function and RFD-Function were comparable (r²=0.15 [0.10:0.20] vs. r²=0.13 [0.09:0.17]; difference=0.03 ‘r² points’ [-0.01:0.06], p=0.158) (Table 2, Figure 2C). Furthermore, associations between early RFD-Function (N=4) and late RFD-Function (N=6) were comparable (r²=0.14 [0.09:0.20] vs. r²=0.14 [0.12:0.16], respectively; difference=0.01 ‘r² points’ [-0.01:0.03], p=0.209) (data not shown).

Discussion

This systematic review is the first to summarize existing evidence on the importance of lower extremity RFD (as well as Fmax) for physical function in neurodegenerative populations, comprising aging and common neurodegenerative disorders (PD, MS, and Stroke). Overall, large heterogeneity was observed regarding deriving/calculating RFD from force-time curves as well as assessment of physical function (balance in particular). Also, most of the identified studies had sample sizes of n ≤40, and only N=8 out of N=32 investigated neurodegenerative patients with none in AD.

The analysis of deficits (Part 1) identified a preferentially impaired ability to generate RFD, compared to Fmax, in aging (individuals having inferior vs. superior physical function) and PD (patients vs. healthy controls). While this was accompanied by limitations in physical function, no associations were observed between these deficits. The analysis of reported associations (Part 2) revealed that Fmax and RFD were almost identical in explaining physical function (15% and 13%, respectively, corresponding to weak associations) in aging, PD, MS, and Stroke.

Deficits in RFD, Fmax and physical function (Part 1)

The background for the present review stems from the statements emphasizing that RFD is a particularly sensitive outcome being able to detect acute and/or chronic adaptations in neuromuscular function (i.e. by adapting earlier and/or more extensively than Fmax) and is of particular importance for physical function[38,40,41]. Indeed, the ability to produce a rapid rise in force seems vital as several daily activities such as preventing a fall after postural perturbation[76], walking fast[77], or stairclimbing[78] are characterized by a limited time to develop force (approximately 30-250 ms), prompting less time than it takes to achieve maximal strength (approximately 300-600 ms)[77,79]. Despite the apparent theoretical importance of lower extremity RFD for physical function, strong supporting evidence has previously been lacking with individual studies revealing divergent findings (see Table 2). To exemplify, in older fallers vs. non-fallers, Pijnappels and colleagues[66] reported deficits in RFD and Fmax of a similar magnitude, Kamo and colleagues (the largest identified study)[56] reported deficits in RFD but not Fmax, Bento and colleagues[49] reported no deficits in neither RFD nor Fmax, whereas Palmer and colleagues[64] reported deficits in early RFD but not in late RFD or Fmax. Study findings on the distinction between early and late RFD – with the former suggested to be specifically important for certain functional tasks requiring very short muscle response time (≤100 ms)[64,66] – are nevertheless also divergent and currently inconclusive[57,49,60,64].

The findings of the present review overall support that RFD is particularly sensitive towards neurodegeneration by adapting earlier and/or more extensively compared to Fmax, as indicated by the deficit point estimates and confidence intervals (RFD vs. Fmax difference=10.9% point [5.5:16.3]; Figure 2A, Table 1). This was based on studies involving older individuals and PD patients, with a preferentially impaired ability to generate RFD (and Fmax) in individuals having inferior vs. superior physical function (i.e. older fallers vs. non-fallers, older low functioning vs. high functioning, old vs. young, PD patients vs. healthy controls). A similar pattern was observed across different muscle groups as well as contraction mode, yet most robustly from knee extensor and isometric contractions (Figure 2B). Moreover, while the numerical differences indicated that early phase RFD (compared to late phase RFD) is also sensitive towards neurodegeneration, the point estimates and confidence intervals did not support this conclusion (early vs. late phase RFD difference=2.8% point [-3.5:9.0]).

As individuals having inferior physical function expectedly have a higher degree of neurodegeneration (PD patients by definition), the observed preferential impairments (i.e. deficits) in RFD appear meaningful. Although few studies have examined the existence of such a “neurodegeneration-related” link, Cruickshank and colleagues[80] reported that striatum (brain area involved in inhibiting and facilitating movement) was degenerated in prodromal Huntington’s disease patients compared to healthy controls, and was strongly associated with plantar flexion RFD_200ms. Also, Yamada and colleagues[81] reported that older fallers were characterised by more global brain atrophy compared to non-fallers. Lastly, Hammond and colleagues[82] reported large deficits in voluntary RFD (but not in Fmax) in PD patients compared to healthy controls, yet not in evoked ‘involuntary’ RFD induced by octet/tetanic electrical muscle stimulation, indicating CNS dysfunction as the main underlying contributing factor. Interestingly, it has been shown that at least 50% of dopaminergic neurons in the Substantia Nigra of PD patients are compromised prior to the appearance of traditional functional symptoms[13]. These different study findings should nevertheless be cautiously interpreted as lifestyle factors (physical activity/exercise participation in particular), may potentially influence the extent of neurodegeneration, impairments in neuromuscular function, and limitations in physical function[83-85].

The present review emphasizes the structural and functional CNS changes known to occur in the outlined neurodegenerative populations[5,9-11] as main determinants of RFD, thus expanding our understanding of RFD being particularly reliant on a well-functioning CNS[38,42]. However, other determinants may also have contributed to the observed deficits in RFD such as reduced muscle size, altered muscle architecture, fast-to-slow transition in muscle fiber type composition, reduced intrinsic muscle fiber contractility, and reduced mechanical properties of tendons and aponeuroses[38]. These changes are all commonly observed with advanced age[86-89] (including PD and Stroke patients due to their often-advanced age), but to some extent also in the younger neurodegenerative patient populations (MS[17] and potentially also Stroke) especially with advanced disease progression. However, as previously proposed in narrative reviews[90-92], neurodegeneration often precede and potentially drive these muscular/tendinous changes of the outlined neurodegenerative populations. While this reassures us that neurodegeneration was the main determinant of the observed deficits in RFD, we cannot exclude additional contributions from these other determinants. Moreover, impairments in cognitive function – also commonly observed in the outlined neurodegenerative populations[93-96] – may indirectly influence RFD. This was examined in an aging study, reporting a gender-dependent association between RFD and some domains of cognitive function (men: memory; women: executive function, attention, language)[97]. The link between cognitive function and RFD should nevertheless be interpreted with caution due to the scarcity of studies.

Associations between RFD or Fmax and physical function (Part 2)

Across all studies, the reported associations between RFD and physical function as well as between Fmax and physical function, were of a comparable magnitude (explaining 13% and 15%, respectively) (Figure 2C, Table 2), corresponding to weak associations. While this means that factors other than RFD (along with Fmax) contributed to determining physical function in the outlined neurodegenerative populations, these associations may also have been ‘contaminated’ by high variability in the extracted outcome measures of neuromuscular function (RFD in particular) and physical function.

The present findings are corroborated by a recent impressive large-scale study (n=1089, age range 26-96 years) from Osawa and colleagues[39]. They investigated the association between RFD or Fmax and physical function (walking, chair rise, performance batteries), and reported r²-values ranging from 0.18-0.51 for RFD-Function and from 0.18-0.54 for Fmax-Function. The greater magnitude of these associations compared to that observed in the present study are likely explained by the very large sample size spanning the entire adult lifespan, that further enabled them to adjust for age, race, and BMI. Despite the apparent comparable magnitude of RFD and Fmax being associated with physical function, the RFD-Function association remained almost unaffected after adjusting for Fmax. This provide novel evidence revealing that RFD and Fmax impact physical function independently of each other[39]. We strongly emphasize to keep this notion in mind, when interpreting the findings of the present study and discussing the importance of RFD vs. Fmax in a clinical context.

Due to the large heterogeneity in lower extremity muscle actions and functional tests across the included studies of the present review, we presented deficits in RFD and Fmax across specific muscles groups and contraction modes only, but did not go further into their associations with physical function. A major challenge of interpreting such associations, is that some tests of physical function preferentially rely on specific muscle groups and/or on specific muscle response times (relating to early RFD, late RFD, or Fmax, respectively). This is furthermore complicated by the involvement and contribution of balance/coordination in determining physical function.

Methodological considerations

The present review has a number of limitations that should be kept in mind when interpreting the findings and using this to design future studies. First, large heterogeneity was present in the approach used to derive/calculate RFD in the identified studies, in line with recent observations by Blazevich and colleagues[98]. This contrasts the fact that recommendations on methodological procedures have been put forward[38]. In order to report this divergent data, we chose to summarize and report one value of lower extremity RFD from each study assuming that this overall represent the rising phase of the force-time curve. Second, some heterogeneity was seen in relation to muscle groups (including single- vs. multi-joint) and contraction modes (dynamic vs. isometric) used to derive RFD and Fmax, yet we deemed it too speculative to go further into details on how this could implicate the transfer to functional performance outcomes. To exemplify, isometric testing appear superior since RFD will not be influenced by changes in the force-length relationship of the muscle during shortening[57,99], whereas dynamic testing - especially when this involve multi-joint movements - may better reflect the dynamic movement required during tasks such as locomotion[70]. In relation to the latter, specific tests of physical function often preferentially rely on specific muscle groups and/or on specific muscle response times (relating to early RFD, late RFD, or Fmax, respectively), but also on the involvement and contribution of balance/coordination. This challenges the interpretation of the present findings but especially our general understanding of associations between RFD (of Fmax) and physical function. Fourth, modest sample sizes characterised the included studies, yielding limited statistical power from individual studies. Fifth, our quantitative analysis included sample size weighted averages only, but not variance. This was chosen since we aimed to report deficit (percentage) values for Part 1 and 2, and since no variance was reported for Part 2. Sixth, as we identified, included, and summarized data across different types of observational and intervention studies (see Table 1 and Supplementary Table 2), we chose not to carry out risk of bias and quality assessment. Mainly since no single tool exists that can embrace such diversity of study types. Seventh, the present study findings were summarized across different study populations known to undergo different types of neurodegeneration, which may have impacted our findings. Figure 2A and 2C nevertheless help elucidate any potential population-specific influence.

Clinical implications and perspectives

Identification of modifiable neuromuscular predictors of physical function/disability is of major relevance to the outlined neurodegenerative populations, particularly as this can help optimize counteractive strategies (e.g. rehabilitation and physical exercise). The fact that lower extremity RFD appeared particularly sensitive towards neurodegeneration, indicate that RFD could serve as a useful indicator of changes in the nerve-muscle interaction. Whether RFD (and Fmax) can also be viewed as a biomarker per se is debatable[100], although some areas within aging research are currently presenting muscle strength as a biomarker of deterioration in physical function and progression in disability[101-103].

As lower extremity RFD and MVC were associated with physical function of a comparable magnitude in the outlined neurodegenerative populations, this may imply that RFD is somewhat redundant to assess in a clinical context. Counteractive strategies should altogether focus on and target Fmax (and perhaps also RFD if being meaningful), along with additional important aspects such as balance/coordination as well as aerobic capacity and endurance. On the other hand, the intriguing study findings by Osawa and colleagues do however support to also assess RFD, as it was shown to impact physical function independently of Fmax[39]. To expand our knowledge on the potential independent impact of RFD and Fmax, respectively, future studies should investigate whether some tests of physical function or specific phases thereof preferentially rely on specific muscle groups and/or on specific muscle response times (relating to early RFD, late RFD, or Fmax, respectively).

Conclusion

The present systematic review provided novel summarized data on lower extremity RFD along with its importance for physical function in neurodegenerative populations (i.e. aging and common neurodegenerative disorders). Overall, the findings reveal that lower extremity RFD is (1) particularly sensitive (i.e. adapts earlier and/or more extensively) towards neurodegeneration, and more so than Fmax, and (2) of importance for physical function but apparently not superior to Fmax. Altogether, RFD could serve as a useful indicator (or biomarker) of changes in neuromuscular function elicited by neurodegeneration.

Authors’ Contributions

SDL, UD, and LGH contributed to the conception of the study and to the development of the search strategy. SDL and LGH conducted the systematic search and completed the acquisition of data, performed the data analysis, and took the lead in writing the manuscript. All the authors discussed the results and contributed to the final manuscript.

Funding

Open access funding provided by Department of Public Health, Aarhus University.