Accepted Manuscript

Muscular strength as a predictor of all-cause mortality in apparently healthy

population: a systematic review and meta-analysis of data from approximately 2
million men and women

Antonio García-Hermoso, PhD, Iván Cavero-Redondo, MSc, Robinson Ramírez-

Vélez, PhD, Jonatan Ruiz, PhD, Francisco B. Ortega, PhD, Duck-Chul Lee, PhD,
Vicente Martínez-Vizcaíno, PhD, MD
PII: S0003-9993(18)30079-0
DOI: 10.1016/j.apmr.2018.01.008
Reference: YAPMR 57142

To appear in: ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION

Received Date: 23 September 2017

Revised Date: 29 December 2017
Accepted Date: 5 January 2018

Please cite this article as: García-Hermoso A, Cavero-Redondo I, Ramírez-Vélez R, Ruiz J, Ortega
FB, Lee D-C, Martínez-Vizcaíno V, Muscular strength as a predictor of all-cause mortality in apparently
healthy population: a systematic review and meta-analysis of data from approximately 2 million men
and women, ARCHIVES OF PHYSICAL MEDICINE AND REHABILITATION (2018), doi: 10.1016/
j.apmr.2018.01.008.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT
Muscular strength as a predictor of all-cause mortality in apparently healthy population:

a systematic review and meta-analysis of data from approximately 2 million men and

women

Running head: Muscular strength and all-cause mortality

PT
Antonio García-Hermoso1, PhD, Iván Cavero-Redondo2, MSc, Robinson Ramírez-

RI
Vélez3, PhD, Jonatan Ruiz4, PhD, Francisco B. Ortega4, PhD, Duck-Chul Lee5, PhD,

SC
Vicente Martínez-Vizcaíno2,6, PhD, MD

U
1 Laboratorio de Ciencias de la Actividad Física, el Deporte y la Salud, Facultad de
AN
Ciencias Médicas, Universidad de Santiago de Chile, USACH, Santiago, Chile; 2
M

Universidad de Castilla-La Mancha. Health and Social Research Center, Cuenca, Spain;

3 Centro de Estudios en Medición de la Actividad Física (CEMA), Escuela de Medicina

D

y Ciencias de la Salud, Universidad del Rosario, Bogotá, D.C, Colombia; 4 PROFITH

TE

“PROmoting FITness and Health through physical activity” research group. Department

of Physical Education and Sport, Faculty of Sport Sciences, University of Granada,

EP

Spain; 5 Department of Kinesiology, College of Human Sciences, Iowa State

University, Ames; 6 Universidad Autónoma de Chile, Facultad de Ciencias de la Salud,

C
AC

Talca, Chile

Correspondence to: Antonio García-Hermoso. Laboratorio de Ciencias de la Actividad

Física, el Deporte y la Salud, Facultad de Ciencias Médicas, Universidad de Santiago de

Chile, USACH, Chile. Avenida Libertador Bernardo O'Higgins nº 3363. Estación

 ACCEPTED MANUSCRIPT
Central. Santiago. Chile; Telephone: (+562) 2 718 00 00. E-mail:

antonio.garcia.h@usach.cl

PROSPERO registration: CRD42016032733.

PT
ACKNOWLEDGMENTS

RI
AGH was funded by “Programa Becas Iberoamérica Jóvenes Profesores Investigadores

2016. Santander Universidades”.

U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
Muscular strength as a predictor of all-cause mortality in apparently healthy population: a

systematic review and meta-analysis of data from approximately 2 million men and women

Running head: Muscular strength and all-cause mortality

PT
RI
U SC
AN
M
D
TE
C EP
AC

1
 ACCEPTED MANUSCRIPT
1 ABSTRACT

2 Objective: The aim of the present systematic review and meta-analysis was to determine the

3 relationship between muscular strength and all-cause mortality risk and to examine the sex-specific

4 impact of muscular strength on all-cause mortality in apparently healthy population.

5 Data Sources: Two authors systematically searched MEDLINE, EMBASE and SPORTDiscus

PT
6 databases and conducted manual searching of reference lists of selected articles.

7 Study Selection: Eligible cohort studies were those that examined the association of muscular

RI
8 strength with all-cause mortality in apparently healthy population. The hazard ratio (HR) estimates

SC
9 with 95% confidence interval (CI) were pooled by using random effects meta-analysis models after

10 assessing heterogeneity across studies.

U
11 Data extraction: Two authors independently extracted data.
AN
12 Data Synthesis: Thirty-eight studies with 1,907,580 participants were included in the meta-analysis.

13 The included studies had a total of 63,087 deaths. Higher levels of handgrip strength were
M

14 associated with a reduced risk of all-cause mortality (HR= 0.69; 95% CI, 0.64–0.74) compared to

15 lower muscular strength, with a slightly stronger association in women (HR= 0.60; 95% CI, 0.51–
D

16 0.69) than men (HR= 0.69; 95% CI, 0.62–0.77) (all p<0.001). Also, adults with higher levels of
TE

17 muscular strength, as assessed by knee extension strength test, had a 14% lower risk of death (HR=
EP

18 0.86: 95% CI, 0.80–0.93; p < 0.001) compared to adults with lower muscular strength.

19 Conclusions: Higher levels of upper- and lower-body muscular strength are associated with a lower
C

20 risk of mortality in adult population, regardless of the age and follow-up period. Muscular strength
AC

21 tests can be easily performed to identify people with a lower muscular strength and, consequently,

22 with an increased risk of mortality.

23 Key Words: Hand Strength; Leg Strength; Death; Muscles

24

25

26

2
 ACCEPTED MANUSCRIPT
27 ABBREVIATIONS

28 HR, hazard ratios

29 PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses

30

31

PT
32

33

RI
34

SC
35

36

U
37
AN
38

39
M

40

41
D

42
TE

43
EP

44

45
C

46
AC

47

48

49

50

51

52

3
 ACCEPTED MANUSCRIPT
53 INTRODUCTION

54 Muscular strength is recognized as a marker of cardiometabolic risk that has been associated with

55 morbidity in adults and elderly people1 and is independently associated with adult metabolic

56 syndrome at long-term.2 Similarly, studies have shown that low handgrip strength is associated with

57 sarcopenia,3 functional limitations and disabilities,4 and considered as a useful marker for frailty in

PT
58 the elderly.5 Several studies have also suggested that lower extremity muscle mass is an important

59 determinant of physical performance in older persons.6,7 Moreover, reduced muscular strength has

RI
60 been associated with an increased risk of mortality in many longitudinal studies.8-11,12 Therefore,

SC
61 muscular strength might be used as a potential predictor of morbidity and mortality in the
10,11
62 population. This emphasizes the importance of accumulating knowledge from studies in

63
U
different contexts to determine the cut-offs for different diseases, since they are not currently
AN
64 available in the literature. In 2015, Volaklis and colleagues12 in a narrative review of

65 epidemiological studies to investigate the role of muscular strength as a predictor of mortality,

M

66 described a strong and inverse association of muscular strength with all-cause mortality; this
D

67 association has also confirmed among people with specific disorders such as cardiovascular disease,
TE

68 peripheral artery disease, cancer, renal failure, chronic obstructive pulmonary disease, rheumatoid

69 arthritis, and patients with critical illness.

EP

70 A meta-analysis published in 201013 suggested that the pooled hazard ratio for mortality
C

71 comparing the weakest with the strongest group of handgrip strength (14 studies and 53,476 adults)
AC

72 was 1.67 (95% CI, 1.45 - 1.93). Apart from this meta-analysis, many studies have added new

73 insights,14-27 but most studies used handgrip strength test. No previous meta-analysis has analyzed

74 the sex-specific impact of muscular strength on all-cause mortality, and there is no information

75 about the role of other muscular strength measures such as knee extension strength. Although

76 consistent evidence has proven that lower limb muscular strength is a predictor of the ability to

77 perform daily living activities and this ability is associated to frailty and mortality,28 the direct

78 relationship between lower limb strength and the risk of mortality is still unclear since previous
4
 ACCEPTED MANUSCRIPT
79 cohort studies have showed inconsistent findings. Because handgrip and lower limb muscular

80 strength tests are easy to perform, non-invasive and inexpensive, it is important to include these

81 tests in clinical settings to better understand the clinical importance of muscular strength for the

82 development of public health guidelines. Therefore, the aims of the present systematic review and

83 meta-analysis were to: i) to determine the relationship between muscular strength and all-cause

PT
84 mortality risk; and ii) examine the sex-specific impact of muscular strength on all-cause mortality in

85 apparently healthy population.

RI
86

SC
87 METHODS

88 A systematic review and meta-analysis was conducted following the guidelines of the

U
89 Cochrane Collaboration. Findings were reported according to the Preferred Reporting Items for
AN
90 Systematic Reviews and Meta-Analyses (PRISMA).29 The review was registered in PROSPERO

91 (registration number: CRD42016032733).

M

92
D

93 Search strategy
TE

94 Two authors (AG-H & RR-V) systematically searched MEDLINE, EMBASE and SPORTDiscus

95 databases until 1 July 2017 (Supplementary Data 1). The following terms were used: "muscles" OR
EP

96 "muscle strength" OR "muscular" OR "strength" AND "mortality" OR "survival rate" OR "cause of

97 death". Only published articles in English language were included in the study. In addition, the
C

98 literature search was supplemented through the manual review of reference lists in the selected
AC

99 articles.

100

101 Selection criteria

102 The a priori inclusion criteria for this meta-analysis were: (i) exposure: muscular strength measured

103 using a validated strength test; (ii) main outcome: all-cause mortality risk assessed with hazard

104 ratios (HR- cox proportional hazards model); (iii) participants: relatively healthy youth and adults

5
 ACCEPTED MANUSCRIPT
105 excluding studies in which all patients had chronic diseases such as diabetes, heart failure,

106 hypertension, peripheral artery disease, chronic obstructive pulmonary disease, cancer and patients

107 with critical illness, i.e. we excluded studies of patient groups; and (iv) study design: prospective

108 cohort studies. Two authors (AG-H & RR-V) independently assessed the electronic search results.

109 When an article title seemed relevant, the abstract was reviewed for eligibility. When more

PT
110 information was required, the full text of the article was retrieved and appraised. Any differences in

111 the assessments between the two authors were discussed and, if necessary, a third author was

RI
112 involved in decision making (IC). Reasons for exclusion of identified articles were recorded in all

SC
113 cases. Finally, when two studies used the same sample, we included the study with longer follow-

114 up.

U
115
AN
116 Data collection process and data items

117 Two authors (AG-H & RR-V) independently extracted data including the first author’s name, year
M

118 of publication, enrollment year, duration of follow-up, study location, sample size, age at baseline

119 examination, HR (and their associated 95% confidence interval [CI] or standard error [SE]),
D

120 adjusted variables, method of muscular strength assessment, and outcome of interest and number of
TE

121 cases.
EP

122

123 Risk of bias in individual studies

C

124 An assessment of risk of bias in selected studies was made using an adjusted format of the
AC

125 Newcastle–Ottawa quality assessment scale by two authors (AG-H & RR-V) independently.30 This

126 scale contains eight items categorized into 3 domains (selection, comparability, and exposure). A

127 star system is used to enable semi-quantitative assessment of study quality; such that the highest-

128 quality studies are awarded a maximum of 1 star per item with the exception of the comparability

129 domain, which allows allocating 2 stars. Thus, the score ranges from 0 to 9 stars.

130

6
 ACCEPTED MANUSCRIPT
131 Summary measures

132 All analyses were carried out using STATA (version 14.0, STATA Corporation, College Station,

133 Tex). HRs with associated 95% CIs from studies for each outcome of interest were extracted (used

134 to estimate the risk for mortality) and pooled HR using random effect (DerSimonian and Laird)

135 models was then calculated. The likelihood approach with random effects was used to better

PT
136 account for the inaccuracy in the estimate of between-study variance.31 When the HR was

137 unavailable, we requested corresponding authors to send us their HR data.

RI
138

SC
139 Synthesis of results

140 The percentage of total variation across the studies due to heterogeneity (Cochran’s Q-statistic)32

U
141 was used to calculate the I2 statistics, considering I2 values of <25%, 25–50%, and >50% as small,
AN
142 medium, and large heterogeneity, respectively.33 Sensitivity analyses were conducted to assess the

143 robustness of the summary estimates in order to determine whether a particular study accounted for
M

144 the heterogeneity. Thus, each study was deleted from the model once in order to analyze the

145 influence of each study on the overall results.

D

146
TE

147 Risk of bias across studies

EP

148 Small-study effects bias was assessed using the extended Egger's test34 and presence of publication

149 bias was investigated graphically by funnel plots.

C

150
AC

151 Additional analysis

152 We examined potential sources of heterogeneity including sex and type of muscular test (handgrip

153 strength and knee extension strength) by stratifying meta-analyses by each of these factors.

154 Random-effects meta-regression analyses were used to separately evaluate whether results were

155 different by mean age of participants at baseline and length of follow-up.35 A p-value of <0.05 was

7
 ACCEPTED MANUSCRIPT
156 considered a threshold for statistical significance. All analyses were based on previously published

157 studies, thus no ethical approval was required.

158

159 RESULTS

160 Study selection

PT
161 The electronic search strategy retrieved 12,974 articles. After removing duplicate references and

162 based on title and abstract, 75 articles were read in full. The reasons for exclusion based on full text

RI
163 were: (i) inappropriate outcome measurement (5 articles); (ii) failed exposure (muscular strength) (8

SC
164 articles); (iii) review article (3 articles); (iv) study population (17 articles); and (v) no effect

165 estimates presented (4 articles). One additional study of Loprinzi et al.36 was excluded, because it

U
17
166 described the same cohort as another article of Buckner et al. Finally, 38 studies met our
AN
167 inclusion criteria and were included in the systematic review and 33 studies were included in the

168 meta-analysis (Supplementary Data 2). Five studies was not included in the analysis due to data was
M

169 not available.

D

170
TE

171 Study characteristics

172 Table 1 summarizes the characteristics of the 38 included studies.14-27,37-60 All of them were
EP

173 prospective observational studies.

41
174 The 38 studies included 1,907,580 participants. Sample sizes ranged from 355 to
C

16
175 1,142,599 participants. Participants included only men,14,16,18,52,58,60 women50 or both sexes15,17-
AC

27,37,38,41 44-46,51,53,55-57
176 and the age ranged from 19 to 84.9 years. Studies were conducted in

177 USA,17,24,38,44,46,51,52,60 Europe,13,14,16,20,23,25,27,37,45,50,55,58 Taiwan,18 Japan,21,26,53,56,57 Ghana,22 and 17

178 other countries.15 The length of follow-up ranged from 21.5 months to 43 years (mean 11.5 years).

179 ***Table 1 about here***

180 Muscular strength measurement

8
 ACCEPTED MANUSCRIPT
181 Most studies (n=36) assessed the muscular strength using the handgrip test.14-16,18-27,37-51,53-60 The

182 participants made three, 15, 18, 20, 23, 27, 37, 39, 42, 43, 45, 48, 50, 51, 58, 59 two, 19, 21, 24, 38, 44, 46, 47, 54, 56, 57 or one

183 trials 22, 25, 26, 41, 53 using the dominant, 20, 23, 24, 26, 27, 44, 47 non-dominant, 58 both hands, 15, 19, 21, 22, 37, 40-
42, 49, 56, 57, 59 14, 18, 25, 38, 45, 50, 51, 53, 60 18, 19, 21-23, 27, 39, 43, 45-49, 54, 56, 57
184 or not specified. The highest or

185 mean 15, 40, 44, 50, 58 value was used for the analyses expressed as kilograms.

PT
186 In the remaining studies, the muscular strength was assessed using several tests. Five studies

187 used an isokinetic dynamometer 17, 46, 56, 57, 59 to assess the knee extension strength. Two studies 14, 16

RI
52
188 assessed the knee extension, handgrip, and elbow flexion strength; Ruiz et al. assessed upper

SC
189 body strength with a one repetition maximum supine bench press, and lower body strength with a

190 one repetition maximum seated leg press. However, this study was not included in the pooled

U
191 analysis because this was the only study that used bench and leg presses. Finally, Guadalupe-Grau
AN
192 et al. 20 measured the upper limb (handgrip and shoulder abduction) and lower limb (knee extensors

193 and hip flexors) maximal voluntary isometric strength in all subjects using a hydraulic hand
M

194 dynamometer and a manual muscle test system. Studies determined low and high muscular strength

195 using percentiles, most of them used quartiles 15, 17, 18, 20-23, 37, 38, 41, 43, 46, 51, 53, 55-60 or tertiles 27, 21, 25,
D

26, 40, 45, 47-50

196 using low and high quartiles and tertiles categories as low and high muscular strength,
TE

197 respectively. Finally, all included studies adjusted for BMI or used normalized handgrip strength
EP

198 per body mass (i.e., handgrip [kg]/body mass [kg]). 24

199
C

200 Primary outcomes

AC

201 The included studies had a total of 63,843 death cases (3.35%).

202

203 Risk of bias within studies

204 All 38 studies met at least seven Newcastle–Ottawa quality assessment scale criteria and were

205 considered to have adequate methodological quality. The average total score was 8.5 from a

206 maximum of nine with a range from seven to nine (Supplementary Data 3).

9
 ACCEPTED MANUSCRIPT
207
208 Meta-analysis

209 Figure 2 depicts the HR forest plots of low (reference) versus high muscular strength measured by

210 handgrip or knee extension strength test. There was substantial heterogeneity across the studies in

211 the HR for all-cause mortality (I2= 83.8%), showing a pooled HR of 0.69 (95% CI, 0.64–0.74; p <

212 0.001). Regarding knee extension, the pooled HR of all-cause mortality for low (reference) versus

PT
213 high muscular strength measured by knee extension test was 0.86 (95% CI, 0.80–0.93; p < 0.001),

RI
214 showing high heterogeneity (I2= 88.5%) (Figure 2).

215 When we analyzed the HR of all-cause mortality by sex, the pooled HR estimates for

SC
216 handgrip were 0.69 (95% CI, 0.62–0.77; p < 0.001) for men (I2= 71.0%) and 0.60 (95% CI, 0.51–

217 0.69; p < 0.001) for women (I2= 39.9%) (Figure 3). Regarding knee extension, these estimates were

U
0.79 (95% CI, 0.66–0.95; p < 0.001) for men (I2= 59.5%) and 0.62 (95% CI, 0.26–1.49; p= 0.163)
AN
218

219 for women (I2 = 82.1%) (Supplementary Data 4).

M

220 Based on meta-regression analyses, there were no significant effects of mean age on the HR

221 estimates for both handgrip strength and knee extension strength (p= 0.423 and p= 0.866,
D

222 respectively). Additionally, that there were no effects of length of follow-up on the HR estimates for
TE

223 knee extension strength (p= 0.623), but there were effects of length of follow-up on the HR

224 estimates for handgrip strength (p= 0.030) (Figure 3).

EP

225
C

226 Publication bias and sensitivity analysis

AC

227 When the impact of individual studies was examined by removing studies from the analysis one at a

228 time, we observed that the pooled HR estimate for low muscular handgrip strength decreases after

229 removing data from the Takata et al 2007 56 (HR, 0.82 [95% CI: 0.80–0.84]) and Takata et al 2012
57
230 (HR, 0.83 [95% CI: 0.80–0.85]) studies. The pooled HR for low muscular knee extension

231 strength also decreases after removing Takata et al 2012 57 (HR, 0.89 [95% CI: 0.88–0.91]).

232 Evidence suggesting publication bias was found using Egger’s test for handgrip strength (p <

233 0.001), but not for knee extension strength (p= 0.057). Also, the funnel plots for the relationships of
10
 ACCEPTED MANUSCRIPT
234 handgrip and knee extension strength with all-cause mortality were asymmetric (Supplementary

235 Data 5 and 6).

236

237 DISCUSSION

238 Our meta-analysis shows that adults with higher muscular strength levels, measured by handgrip

PT
239 strength test, had a 31% reduced all-cause mortality risk (HR=0.69, 95% CI 0.64–0.74) compared to

240 those adults with lower muscular strength, with a slightly stronger association and lower

RI
241 heterogeneity in women than men. Similarly, adults with higher knee extension strength levels had

SC
242 a 14% lower risk of death (HR=0.86, 95% CI 0.80–0.93) than adults with lower strength levels.

243 Therefore, low muscle strength levels should be considered as a risk factor for all-cause mortality in

U
244 adults.
AN
245 Our meta-analysis shows that lower levels of handgrip strength increase the all-cause

246 mortality risk, supporting the growing body of literature that have linked muscular strength and risk
M

247 of mortality.13 However, in our study, the figures for heterogeneity statistics were large. At least a

248 substantial part of the differences across studies might be explained by the differences in socio-
D

15, 61
249 economic status among populations , differences in muscular test protocols (repetitions, hand
TE

250 used, etc.) and dynamometers included in the test. Furthermore, the observed associations and
EP

251 heterogeneity could be explained by confounding factors such as comorbidities other than diabetes,

252 cardiovascular disease or by unknown comorbidities. However, attenuated but significant

C

253 associations were also found after adjustment for the above-mentioned factors,14, 15, 18, 19, 21, 23, 25, 27,
AC

38, 41-43, 45-47, 49, 50, 52, 54, 56, 57, 59

254 therefore confounding cannot fully account for the observed

255 associations.

256 The underlying mechanisms are not fully understood, but some possible explanations have

257 been proposed. It has been hypothesized that the negative association between muscular strength

258 and chronic inflammation is a potential mechanism behind the favorable influence of muscular
62, 63
259 strength on health. Another possible mechanism through which poor muscular strength may

11
 ACCEPTED MANUSCRIPT
260 influence health includes an increase in the lipoprotein lipase enzyme, which is related to increased

261 accumulation of triglycerides and decreased high-density lipoprotein cholesterol. 64, 65

262

263 Various studies have analyzed the effect of sex on the association between muscular strength
20, 24, 38, 42, 43, 66, 67
264 and all-cause mortality showing inconsistent findings. Our results suggest the

PT
265 influence of high muscular strength on mortality was slightly higher in women than in men. 20, 27, 38,
67
266 These findings are in line with those reported by Guadalupe-Grau et al. 20 showing that the worst

RI
267 muscular strength quartiles (i.e. 1st quartile) of the four measured sites (grip, shoulder, knee and hip)

SC
268 were more closely associated with higher mortality risk in women (HR=4.39) than in men

269 (HR=2.46). Recently, Peterson and colleges 24 found that the predictive power of baseline handgrip

U
270 strength on survival (all-cause mortality) was stronger for Mexican American women than men,
AN
271 suggesting that weakness or declines in muscular strength, relative to body mass, implies greater

272 risk of mortality for women. Potential explanations for the sex-based differences could be the
M

273 following: (i) it has been also hypothesized that, because women have lower values of absolute

274 muscle strength than men, they may be closer to a disability threshold; 20 (ii) sex-specific hormones
D

275 are another possible explanations, for example, the triglyceride levels and blood pressure were
TE

276 lower among women compared to men;68 (iii) it is also plausible that weakness or declines in
EP

277 muscular strength, relative to body mass, poses greater risk of mortality for women than men; 24 and

278 (iv) it has been observed that postmenopausal estrogen levels are directly associated with
C

69
279 cardiovascular and cancer risk,70 and thus, the protective effects of high strength levels may be
AC

280 somewhat masked by the protective effects caused by postmenopausal reductions in estrogen levels.

281 Because muscular strength may decline at different rates depending on the muscle groups
71
282 over time, it seems necessary to measure strength at different sites, especially at the lower and

283 upper limbs. Although lower limb strength is deemed to be of great importance for daily life (i.e.

284 legs for the ability to walk), the magnitude of its association with mortality risk remains

285 controversial. Strength losses in lower limb muscles determine the whole body functionality and

12
 ACCEPTED MANUSCRIPT 20
286 may prevail over upper limb muscular strength in its relationship with mortality. Our meta-

287 analysis shows that stronger adults, as assessed by knee extension strength test, had a 14%

288 (HR=0.86, 95% CI 0.80–0.93) lower risk of death than the weaker individuals. Overall, adults with

289 low muscular strength have more difficulties to perform daily living activities72 and, as

290 consequence, their physical activity levels are decreased, leading to a greater muscle mass losses

PT
73
291 (sarcopenia), which is characterized by a decrease in contractile protein content and excessive

292 intra- and extra-cellular lipid accumulation. 74 This could make adults more vulnerable to accidents,

RI
293 such as injurious falls, or other adverse events; therefore their recovery from acute diseases, injury,

SC
294 or surgery may be compromised.

295

U
296 Study Limitations
AN
297 Our study includes data from approximately two million adults with a mean follow-up greater than

298 10 years, and from various countries, ethnic origins, and socioeconomic backgrounds. However,
M

299 there are also some limitations that should be considered when interpreting the results. First of all,

300 although the included studies have prospective follow up designs, these observational studies
D

301 prevent us from making strong conclusions on the causal role of muscular strength and mortality.
TE

302 Second, although the potential confounders are statistically adjusted in all studies, we could not
EP

303 exclude the possibility that residual confounding underlies the association between muscular

304 strength and mortality in this meta-analysis. Indeed, we were unable to determine whether other
C

305 competing risks or unmeasured confounding (i.e., other lifestyle risk factors, comorbidities or
AC

306 socioeconomic status) may have influenced the observed association of muscular strength with

307 mortality.24 Third, non-Hispanic white adults were predominantly recruited in most studies; thus it

308 is important to examine whether race or ethnicity may affect the association between muscular

309 strength and mortality. To fully confirm this observation, further studies in diverse race and ethnic

310 populations are needed. Fourth, our exploratory analyses could not fully explain the significant

311 heterogeneity in the analyses. Finally, the categorization of muscular strength in the studies was

13
 ACCEPTED MANUSCRIPT
312 heterogeneous and may have led to overestimation of the reported associations; also differences in

313 sex distribution among studies may be behind discrepancies regarding all-cause mortality.

314 CONCLUSION

315 In conclusion, handgrip and knee extension strength were an independent predictor of all-cause

316 mortality in apparently healthy population. Therefore, muscular strength might be considered as a

PT
317 biomarker of ageing and mortality75. For example, Muscular strength, assessed by the handgrip

318 strength test, could be easily and universally applied to identify frail people at increased risk of

RI
319 premature mortality.

SC
320

321 DISCLOSURES

U
322 None
AN
323

324 REFERENCES
M

325 1 Silventoinen K, Magnusson PK, Tynelius P, Batty GD, Rasmussen F. Association of body size
D

326 and muscle strength with incidence of coronary heart disease and cerebrovascular diseases: a
TE

327 population-based cohort study of one million Swedish men. Int J Epidemiol 2009; 38: 110-8.

328 2 Fraser BJ, Huynh QL, Schmidt MD, Dwyer T, Venn AJ, Magnussen CG. Childhood muscular
EP

329 fitness phenotypes and adult metabolic syndrome. Med Sci Sports Exerc 2016;48:1715-22

330 3 Manini TM, Clark BC. Dynapenia and aging: an update. J Gerontol A Biol Sci Med Sci 2012;
C

331 67:28-40
AC

332 4 Rantanen T, Guralnik JM, Foley D, Masaki K, Leveille S, Curb JD, White L. Midlife hand grip

333 strength as a predictor of old age disability. JAMA 1999; 281:558-60.

334 5 Takahashi T, Sugie M, Nara M, Koyama T, Obuchi SP, Harada K, Kyo S, Ito H. Femoral muscle

335 mass relates to physical frailty components in community dwelling older people. Geriatr Gerontol

336 Int 2017; doi: 10.1111/ggi.12945.

14
 ACCEPTED MANUSCRIPT
337 6 Reid KF, Naumova EN, Carabello RJ, Phillips EM, Fielding RA. Lower extremity muscle mass

338 predicts functional performance in mobility-limited elders. J Nutr Health Aging 2008; 12:493-8.

339 7 Visser M, Deeg DJ, Lips P, Harris TB, Bouter LM. Skeletal muscle mass and muscle strength in

340 relation to lower extremity performance in older men and women. J Am Geriatr Soc 2000; 48:381-

341 6.

PT
342 8 Kamiya K, Masuda T, Tanaka S, Hamazaki N, Matsue Y, Mezzani A, et al. Quadriceps strength

343 as a predictor of mortality in coronary artery disease. Am J Med 2015; 128: 1212-9.

RI
344 9 Larsen BA, Wassel CL, Kritchevsky SB, et al. Association of Muscle Mass, Area, and Strength

SC
345 with Incident Diabetes in Older Adults: The Health ABC Study. J Clin Endocrinol Metab 2016;

346 101:1847-55.

U
347 10 Leon AS, Franklin BA, Costa F, et al. Cardiac rehabilitation and secondary prevention of
AN
348 coronary heart disease an american heart association scientific statement from the council on

349 clinical cardiology (subcommittee on exercise, cardiac rehabilitation, and prevention) and the
M

350 council on nutrition, physical activity, and metabolism (subcommittee on physical activity), in

351 collaboration with the american association of cardiovascular and pulmonary rehabilitation.
D

352 Circulation 2005; 111: 369-76.

TE

353 11 Peterson MD, Krishnan C. Growth Charts for Muscular Strength Capacity With Quantile
EP

354 Regression. Am J Prev Med 2015; 49: 935-8.

355 12 Volaklis KA, Halle M, Meisinger C. Muscular strength as a strong predictor of mortality: A
C

356 narrative review. Eur J Intern Med 2015; 26: 303-10.

AC

357 13 Cooper R, Kuh D, Hardy R. Objectively measured physical capability levels and mortality:

358 systematic review and meta-analysis. BMJ 2010; 341: c4467.

359 14 Timpka S, Petersson IF, Zhou C, Englund M. Muscle strength in adolescent men and risk of

360 cardiovascular disease events and mortality in middle age: a prospective cohort study. BMC Med

361 2014; 12: 1. doi: 10.1186/1741-7015-12-62.

15
 ACCEPTED MANUSCRIPT
362 15 Leong DP, Teo KK, Rangarajan S, Lopez-Jaramillo P, et al. Prognostic value of grip strength:

363 findings from the Prospective Urban Rural Epidemiology (PURE) study. Lancet 2015; 386: 266-73.

364 16 Ortega FB, Silventoinen K, Tynelius P, Rasmussen F. Muscular strength in male adolescents and

365 premature death: cohort study of one million participants. BMJ 2012; 345: e7279. doi:

366 10.1136/bmj.e7279.

PT
367 17 Buckner SL, Loenneke JP, Loprinzi PD. Lower extremity strength, systemic inflammation and

368 all-cause mortality: application to the “fat but fit” paradigm using cross-sectional and longitudinal

RI
369 designs. Physiol Behav 2015; 149: 199-202.

SC
370 18 Chen P-J, Lin M-H, Peng L-N, et al. Predicting cause-specific mortality of older men living in

371 the Veterans home by handgrip strength and walking speed: a 3-year, prospective cohort study in

U
372 Taiwan. J Am Med Dir Assoc 2012; 13: 517-21.
AN
373 19 Cooper R, Strand BH, Hardy R, Patel KV, Kuh D. Physical capability in mid-life and survival

374 over 13 years of follow-up: British birth cohort study. BMJ 2014; 348: g2219. doi:
M

375 10.1136/bmj.g2219.

376 20 Guadalupe-Grau A, Carnicero JA, Gómez-Cabello A, et al. Association of regional muscle

D

377 strength with mortality and hospitalisation in older people. Age Ageing 2015; 44: 790-5.
TE

378 21 Kishimoto H, Hata J, Ninomiya T, et al. Midlife and late-life handgrip strength and risk of
EP

379 cause-specific death in a general Japanese population: the Hisayama Study. J Epidemiol Community

380 Health 2014; 68: 663-8.

C

381 22 Koopman JJ, van Bodegom D, van Heemst D, Westendorp RG. Handgrip strength, ageing and
AC

382 mortality in rural Africa. Age Ageing 2015; 44:465-70.

383 23 Legrand D, Vaes B, Matheï C, Adriaensen W, Van Pottelbergh G, Degryse JM. Muscle strength

384 and physical performance as predictors of mortality, hospitalization, and disability in the oldest old.

385 J Am Geriatr Soc 2014; 62: 1030-8.

16
 ACCEPTED MANUSCRIPT
386 24 Peterson MD, Zhang P, Duchowny KA, Markides KS, Ottenbacher KJ, Al Snih S. Declines in

387 Strength and Mortality Risk Among Older Mexican Americans: Joint Modeling of Survival and

388 Longitudinal Data. J Gerontol A Biol Sci Med Sci 2016; 71:1646-52.

389 25 Stenholm S, Mehta NK, Elo IT, Heliövaara M, Koskinen S, Aromaa A. Obesity and muscle

390 strength as long-term determinants of all-cause mortality—a 33-year follow-up of the Mini-Finland

PT
391 Health Examination Survey. Int J Obes 2014; 38: 1126-32.

392 26 Taniguchi Y, Fujiwara Y, Murayama H, et al. Prospective Study of Trajectories of Physical

RI
393 Performance and Mortality Among Community-Dwelling Older Japanese. J Gerontol A Biol Sci

SC
394 Med Sci 2016; 71: 1492-9.

395 27 Arvandi M, Strasser B, Meisinger C, et al. Gender differences in the association between grip

U
396 strength and mortality in older adults: results from the KORA-age study. BMC Geriatr 2016; 16:
AN
397 201.

398 28 Xue QL, Beamer BA, Chaves PH, et al. Heterogeneity in rate of decline in grip, hip, and knee
M

399 strength and the risk of all‐cause mortality: the women's health and aging study II. J Am Geriatr
D

400 Soc. 2010;58(11):2076-84.

TE

401 29 Liberati A, Altman DG, Tetzlaff J, et al. The PRISMA statement for reporting systematic

402 reviews and meta-analyses of studies that evaluate health care interventions: explanation and
EP

403 elaboration. Ann Intern Med 2009; 151: 65-94.

404 30 Wells G, Shea B, O’connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing the
C

405 quality of nonrandomised studies in meta-analyses. Department of Epidemiology and Community

AC

406 Medicine, University of Ottawa, Canada. Retrieved from:

407 www.ohri.ca/programs/clinical_epidemiology/oxford.asp.

408 31 Hardy RJ, Thompson SG. A likelihood approach to meta-analysis with random effects. Stat Med

409 1996; 15: 619-29.

410 32 Higgins J, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses.

411 BMJ 2003; 327: 557-60.

17
 ACCEPTED MANUSCRIPT
412 33 Higgins J, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med 2002; 21:

413 1539-58.

414 34 Egger M, Smith GD, Schneider M, Minder C. Bias in meta-analysis detected by a simple,

415 graphical test. BMJ 1997; 315: 629-34.

416 35 Thompson SG, Sharp SJ. Explaining heterogeneity in meta-analysis: a comparison of methods.

PT
417 Stat Med 1999; 18: 2693-708.

418 36 Loprinzi PD. Lower extremity muscular strength, sedentary behavior, and mortality. Age 2016;

RI
419 38: 1-5.

SC
420 37 Ahmad R, Bath PA. Identification of risk factors for 15-year mortality among community-

421 dwelling older people using Cox regression and a genetic algorithm. J Gerontol A Biol Sci Med Sci

422 2005; 60:1052-8.

U
AN
423 38 Al Snih S, Markides KS, Ray L, Ostir GV, Goodwin JS. Handgrip strength and mortality in

424 older Mexican Americans. J Am Geriatr Soc 2002; 50:1250-6.

M

425 39 Anstey KJ, Luszcz MA, Giles LC, et al. Demographic, health, cognitive, and sensory variables
D

426 as predictors of mortality in very old adults. Psychol Aging 2001;16(1):3-11.

TE

427 40 Celis-Morales C, Lyall DM, Anderson J. The association between physical activity and risk of

428 mortality is modulated by grip strength and cardiorespiratory fitness: evidence from 498 135 UK-
EP

429 Biobank participants. Eur Heart J 2017 38(2):116-22.

430 41 Cesari M, Onder G, Zamboni V, et al. Physical function and self-rated health status as predictors
C

431 of mortality: results from longitudinal analysis in the ilSIRENTE study. BMC Geriatr 2008; 8: 34.
AC

432 doi: 10.1186/1471-2318-8-34.

433 42 Gale CR, Martyn CN, Cooper C, et al. Grip strength, body composition, and mortality. Int J

434 Epidemiol 2007;36(1):228-35.

435 43 Katzmarzyk PT, Craig CL. Musculoskeletal fitness and risk of mortality. Med Sci Sports Exerc

436 2002;34(5):740-44.

18
 ACCEPTED MANUSCRIPT
437 44 Klein BE, Klein R, Knudtson MD, Lee KE. Frailty, morbidity and survival. Arch Gerontol

438 Geriatr 2005; 41: 141-9.

439 45 Ling CH, Taekema D, de Craen AJ, Gussekloo J, Westendorp RG, Maier AB. Handgrip strength

440 and mortality in the oldest old population: the Leiden 85-plus study. CMAJ 2010; 182: 429-35.

441 46 Newman AB, Kupelian V, Visser M, et al. Strength, but not muscle mass, is associated with

PT
442 mortality in the health, aging and body composition study cohort. J Gerontol A Biol Sci Med Sci

443 2006; 61: 72-7.

RI
444 47 Nofuji Y, Shinkai S, Taniguchi Y, et al. Associations of walking speed, grip strength, and

SC
445 standing balance with total and cause-specific mortality in a general population of Japanese elders.

446 J Am Med Dir Assoc 2016;17(2):184.

U
447 48 Rantanen T, Harris T, Leveille SG, et al. Muscle strength and body mass index as long-term
AN
448 predictors of mortality in initially healthy men. J Gerontol A Biol Sci Med Sci 2000;55(3):168-73.

449 49 Rantanen T, Volpato S, Ferrucci L, et al. Handgrip Strength and Cause Specific and Total
M

450 Mortality in Older Disabled Women: Exploring the Mechanism. J Am Geriatr Soc 2003;51(5):636-

451 41.
D

452 50 Rolland Y, Lauwers-Cances V, Cesari M, Vellas B, Pahor M, Grandjean H. Physical

TE

453 performance measures as predictors of mortality in a cohort of community-dwelling older French

EP

454 women. Eur J Epidemiol 2006; 21: 113-22.

455 51 Rothman MD, Leo-Summers L, Gill TM. Prognostic significance of potential frailty criteria. J
C

456 Am Geriatr Soc 2008; 56: 2111-6.

AC

457 52 Ruiz JR, Sui X, Lobelo F, et al. Association between muscular strength and mortality in men:

458 prospective cohort study. BMJ 2008;337:a439.

459 52 Shibata H, Haga H, Nagai H, et al. Predictors of all-cause mortality between ages 70 and 80: the

460 Koganei study. Arch Gerontol Geriatr 1992; 14: 283-97.

19
 ACCEPTED MANUSCRIPT
461 54 Strand BH, Cooper R, Bergland A, et al. The association of grip strength from midlife onwards

462 with all-cause and cause-specific mortality over 17 years of follow-up in the Tromsø Study. J

463 Epidemiol Community Health 2016;70(12):1214-21.

464 55 Syddall H, Cooper C, Martin F, Briggs R, Sayer AA. Is grip strength a useful single marker of

465 frailty? Age Ageing 2003; 32: 650-6.

PT
466 56 Takata Y, Ansai T, Akifusa S, et al. Physical fitness and 4-year mortality in an 80-year-old

467 population. J Gerontol A Biol Sci Med Sci 2007; 62: 851-8.

RI
468 57 Takata Y, Shimada M, Ansai T, et al. Physical performance and 10-year mortality in a 70-year-

SC
469 old community-dwelling population. Aging Clin Exp Res 2012; 24: 257-64.

470 58 Van den Beld AW, Visser TJ, Feelders RA, Grobbee DE, Lamberts SW. Thyroid hormone

U
471 concentrations, disease, physical function, and mortality in elderly men. J Clin Endocrinol Metab
AN
472 2005; 90: 6403-9.

473 59 Veronese N, Stubbs B, Fontana L, et al. A comparison of objective physical performance tests
M

474 and future mortality in the elderly people. J Gerontol A Biol Sci Med Sci 2016;72(3):362-68.

475 60 Willcox BJ, He Q, Chen R, et al. Midlife risk factors and healthy survival in men. JAMA 2006;
D

476 296: 2343-50.

TE

477 61 Stenholm S, Maggio M, Lauretani F, et al. Anabolic and catabolic biomarkers as predictors of
EP

478 muscle strength decline: the InCHIANTI study. Rejuvenation Res 2010; 13: 3-11.

479 62 Jurca R, Lamonte MJ, Church TS, et al. Associations of muscle strength and aerobic fitness with
C

480 metabolic syndrome in men. Med Sci Sports Exerc 2004; 36: 1301-7.
AC

481 63 Sente T, Van Berendoncks AM, Hoymans VY, Vrints CJ. Adiponectin resistance in skeletal

482 muscle: pathophysiological implications in chronic heart failure. J Cachexia Sarcopenia Muscle

483 2016; 7: 261-74.

484 64 Coles C. Adipokines in Healthy Skeletal Muscle and Metabolic Disease. In: White, J, Smythe,

485 G, eds. Growth Factors and Cytokines in Skeletal Muscle Development, Growth, Regeneration and

486 Disease. Switzerland: Springer, 2016: 133-60.

20
 ACCEPTED MANUSCRIPT
487 65 Freitas D, Maia J, Beunen G, et al. Socio-economic status, growth, physical activity and fitness:

488 the Madeira Growth Study. Ann Hum Biol 2007; 34: 107-22.

489 66 Fujita Y, Nakamura Y, Hiraoka J, et al. Physical-strength tests and mortality among visitors to

490 health-promotion centers in Japan. J Clin Epidemiol 1995; 48: 1349-59.

491 67 Sasaki H, Kasagi F, Yamada M, Fujita S. Grip strength predicts cause-specific mortality in

PT
492 middle-aged and elderly persons. Am J Med 2007; 120: 337-42.

493 68 Lew J, Sanghavi M, Ayers CR, McGuire DK, Omland T, Atzler D, Gore MO, Neeland I, Berry

RI
494 JD, Khera A, Rohatgi A, de Lemos JA. Sex-based differences in cardiometabolic biomarkers.

SC
495 Circulation 2017; 135:544-55

496 69 Pérez-López FR, Larrad-Mur L, Kallen A, Chedraui P, Taylor HS. Gender differences in

U
497 cardiovascular disease: hormonal and biochemical influences. Reprod Sci 2010; 17:511-31
AN
498 70 Key TJ, Appleby PN, Reeves GK, Roddam AW, et al. Circulating sex hormones and breast

499 cancer risk factors in postmenopausal women: reanalysis of 13 studies. Br J Cancer 2011; 105:709-
M

500 22.

501 71 Frontera WR, Hughes VA, Fielding RA, Fiatarone MA, Evans WJ, Roubenoff R. Aging of
D

502 skeletal muscle: a 12-yr longitudinal study. J Appl Physiol 2000; 88: 1321-6.
TE

503 72 Rantanen T, Guralnik JM, Sakari-Rantala R, et al. Disability, physical activity, and muscle
EP

504 strength in older women: the Women's Health and Aging Study. Arch Phys Med Rehabil 1999; 80:

505 130-5.
C

506 73 Roubenoff R. Sarcopenia and its implications for the elderly. Eur J Clin Nutr 2000; 54: S40-47.
AC

507 74 Kent-Braun JA, Ng AV, Young K. Skeletal muscle contractile and noncontractile components in

508 young and older women and men. J Appl Physiol 2000; 88: 662-8.

509 75 Sayer AA, Kirkwood T. Grip strength and mortality: a biomarker of ageing? Lancet (London,

510 England) 2015;386(9990):226.

511

512

21
 ACCEPTED MANUSCRIPT
513 FIGURE LEGENDS

514 Figure 1. Hazard ratios of all-cause mortality comparing low (reference) versus high muscular

515 strength measured by handgrip and knee extension strength. a, men only; b, women only.

516 Figure 2. Hazard ratios of all-cause mortality comparing low (reference) versus high muscular by

517 sex.

PT
518 Figure 3. Associations of mean age (years) and length of follow-up (years) with hazard ratios of all-

519 cause mortality by muscular strength test. Solid line indicates a linear relationship. Size of each data

RI
520 point is proportional to its statistical weight.

U SC
AN
M
D
TE
C EP
AC

22

Table 1. Characteristics of included studies.
Study, year (reference)
Ahmad and Bath 2005
Al Snih et al. 2002
Anstey et al. 2001
Arvandi et al. 2016
Buckner et al. 2015
Celis-Morales et al. 2017
Cesari et al. 2008
ACCEPTED MANUSCRIPT
Age, y
Study name, country and race Sample size Men, % Follow-up, y Muscle strength test Adjustment for Covariates Number of deaths
(mean or range)
Nottingham Longitudinal 1042 75.2 40.5 15 Handgrip strength NA 741
Study of Activity and Ageing, UK
PT
Hispanic Established Population for 2488 72.8 43.5 5 Handgrip strength Socioeconomic status, 307
the Epidemiological Study of the functional disability, timed
Elderly, USA walked, medical conditions,
RI
Mexican Americans BMI, smoking
Australian Longitudinal Study of 1947 79,6 53.3 5.76 Handgrip strength Age, gender, education and 680
SC
Ageing self-rated health
The Cooperative Health Research in 1066 76.0 49.7 3 Handgrip strength Age, gender, adjusted for age, 95
the Region of Augsburg (KORA)- nutritional status, physical
U
Age Study inactivity, number of
prescribed drugs, lung
diseases, cardiovascular
AN
disease and cancer
NHANES, USA 1363 62.4 47.6 3 Knee extension strength Age, weight status, 118
M
Mexican Americans (3.4%) medication use, gender, race-
Other Hispanic (5.1%) ethnicity, MVPA, total
Non-Hispanic White (81.3%) cholesterol, diabetes, mean
D
Non-Hispanic Black (7.1%) arterial pressure, arthritis,
coronary artery disease,
TE
stroke, smoking status and
use of ambulatory devise
UK-Biobank 498 135 56.55 43.6 4.9 Handgrip strength Age, gender, ethnicity, 8591
EP
Whites (94.6%) deprivation index, BMI,
South Asians (2.0%) smoking status, total
Blacks (1.6%) sedentary time, alcohol
Chinese (0.3%) intake, depression, stroke,
C
Mixed background (0.6%) angina, heart attack,
Others (0.9%) hypertension, cancer,
diabetes, or long-standing
AC
illness
Invecchiamento e Longevità nel 335 85.6 43.0 2 Handgrip strength Age, gender, BMI, cognitive 71
Sirente study, USA performance scale, number of
clinical conditions, albumin,
total cholesterol, C-reactive
protein
 ACCEPTED MANUSCRIPT

Age, y
Study, year (reference) Study name, country and race Sample size Men, % Follow-up, y Muscle strength test Adjustment for Covariates Number of deaths
(mean or range)

Chen et al. 2012 China 558 82.4 100 3 Handgrip strength Age, height, body mass 99
index, waist circumference,
smoking habits, physical
activity, hypertension,

PT
diabetes mellitus, proteinuria,
walking speed, hemoglobin,
platelet, neutrophil count,

RI
lymphocyte count, HbA1C,
alanine transaminase,
creatinine, albumin, uric acid,
serum triglyceride, serum

SC
total cholesterol, serum high-
density lipoprotein

Cooper et al. 2014 The Medical Research Council 2984 53.0 45.4 13 Handgrip strength Gender, height, BMI, 177

U
National Survey of Health and socioeconomic position,
Development, England, Scotland and smoking, physical activity,

AN
Wales and health status

Gale et al. 2006 UK Department of Health and Social 800 74.5 56.5 24 Handgrip strength Age, gender and body size 756

M
Security

D
Guadalupe-Grau et al. Toledo Study for Healthy Aging, 1755 ≥65 43.9 5.5 Upper limb and lower Age, BMI, educational level, 287

TE
2015 Spain limb maximal voluntary Geriatric depression Scale
isometric strength, and and Charlson Index
handgrip strength
EP
Katzmarzyk and Craig, Canada Fitness Survey 8116 35.6 48.4 13 Handgrip strength Age, smoking status, body 238
2002 mass, and VO2max

Kishimoto et al. 2014 Prospective study of cardiovascular 2527 ≥40 42.1 19 Handgrip strength Age, systolic blood pressure, 783
C

and malignant diseases, Japan use of antihypertensive

agents, diabetes, total
AC

cholesterol, BMI,
electrocardiogram
abnormalities, smoking,
alcohol intake and physical
activity

Klein et al. 2005 Beaver Dam Eye Study Cohort, USA 2962 68.9 43.0 3.5 Handgrip strength Age, gender, hypertension, 149
diabetes, and history of
cardiovascular disease
 ACCEPTED MANUSCRIPT

Age, y
Study, year (reference) Study name, country and race Sample size Men, % Follow-up, y Muscle strength test Adjustment for Covariates Number of deaths
(mean or range)

Koopman et al. 2015 Ghana 923 ≥50 52.0 2 Handgrip strength Age, gender, and BMI 46

Legrand et al. 2014 Belgium 560 84.9 36.6 1 and 9.5 Handgrip strength Gender, individual 129

PT
months morbidities, muscle mass,
and serum inflammatory
marker levels

RI
Leong et al. 2015 Prospective Urban Rural 139 691 35-70 41.9 4 Handgrip strength Age, gender, education, 3379
Epidemiology study, 17 countries employment status, physical
South Asian (21%) activity, tobacco and alcohol

SC
Chinese (33%) use, daily dietary energy
Malaysian (7%) intake, proportion of caloric
Persian (5%) intake from protein, self-
Arab (1%) reported hypertension,
African (3%) diabetes, heart failure,

U
European (11%) coronary artery disease, and
Latin American (16%) chronic obstructive

AN
Other (3%) pulmonary disease, and self-
reported prior stroke or
cancer, BMI, and waist-to-hip
ratio

M
Ling et al. 2010 Population-based Leiden 555 ≥85 35.0 9.5 Handgrip strength Comorbidities, total number 444
85-plus study, The Netherlands of prescription medications

D
and smoking, body surface
area at age 89 years, and
absolute change in scores on

TE
the Mini-Mental State
Examination, Geriatric
Depression Scale, Activities
of Daily Living disability
EP
scale, Instrumental Activities
of Daily Living disability
scale, and level of physical
activity over four years
C

Newman et al. 2006 Health ABC Study, USA 2292 73.7 49.0 4.9 Knee extension and Age, race, height, smoking, 286
AC

Black (38.7%) handgrip strength physical activity, number of

White (61.3%) chronic conditions, education,
interleukin-6, and Center for
Epidemiologic Studies-
Depression scale score
 ACCEPTED MANUSCRIPT

Age, y
Study, year (reference) Study name, country and race Sample size Men, % Follow-up, y Muscle strength test Adjustment for Covariates Number of deaths
(mean or range)

Nofuji et al. 2016 Tokyo Metropolitan Institute of 1085 72 42.5 10.3 Handgrip strength Age, gender, study area, 324
Gerontology Longitudinal education, body mass index,
Interdisciplinary Study on Aging stroke, heart disease,
hypertension, diabetes

PT
mellitus, cystatin C, IL-6,
high-sensitivity c-reactive
protein, albumin,

RI
hemoglobin, total cholesterol,
self-rated health, depressive
mood, smoking, alcohol,
physical activity, walking

SC
speed and standing balance

Ortega et al. 2012 Swedish military conscription 1 142 599 16-19 100 24.2 Knee extension and Birth cohort, conscription age 26145
register for male adolescents, handgrip strength and conscription office

U
Sweden

AN
Peterson et al. 2016 Hispanic Established Population for 3050 75.7 42.3 18 Handgrip strength Age, sex, diabetes, disability 2178
the Epidemiological Study of status, and various
the Elderly, USA sociodemographic
Mexican Americans variables

M
Rantanen et al. 2000 Honolulu Heart Program 6040 54 100 30 Handgrip strength Age, education, occupation, 2900
smoking, leisure time

D
physical activity and body
height

TE
Rantanen et al. 2003 Women’s Health and Aging Study 6521 78.3 0 5 Handgrip strength Age, weight, height, race, 336
congestive heart failure,
stroke, chronic obstructive
pulmonary disease, diabetes
EP
mellitus, cancer, hand
osteoarthritis at baseline,
smoking, walking, Geriatric
Depression Scale, serum
C

albumin, weight loss, self-

reported unintentional loss of
AC

weight, C-reactive protein

and interleukin-6
 ACCEPTED MANUSCRIPT

Age, y
Study, year (reference) Study name, country and race Sample size Men, % Follow-up, y Muscle strength test Adjustment for Covariates Number of deaths
(mean or range)

Rolland et al. 2006 EPIDOS study, France 7250 80.5 0 4 Handgrip strength Physical performance score, 722
diabetes, cancer, Pfeiffer
cognitive test, smoking, self-
reported assessment of health,

PT
obesity, inability to walk
outdoors, and hospitalization
during the year

RI
Rothman et al. 2008 Precipitating Events Project, USA 754 78.4 35.4 7.5 Handgrip strength Age, sex, race, education, 283
Non-Hispanic white (90.5%) chronic conditions, and the
presence of the other six

SC
frailty criteria

Ruiz et al. 2008 Aerobics centre longitudinal study, 8762 42.3 100 18.9 Maximal leg and bench Age, physical activity, 503
USA press strength smoking, alcohol intake,

U
BMI, baseline medical
conditions, family history of

AN
cardiovascular disease, and
cardiorespiratory fitness

Shibata et al. 1992 The Koganei Study, Japan 422 53.3 47.7 10 Handgrip strength NA 102

M
Stenholm et al. 2014 Mini-Finland Health Examination 3594 ≥50 40.8 17.9 Handgrip strength Age, gender, education, 3043
Survey, Finland smoking, alcohol use,

D
physical activity, BMI,
baseline hypertension,

TE
cardiovascular disease,
diabetes and cancer

Strand et al. 2016 Tromsø Study 6850 62 41.7 17 Handgrip strength Adjusted for age, gender, 2338
EP
BMI, height, systolic blood
pressure, total cholesterol,
triglycerides, self-reported
general health status, self-
C

reported history of heart

attack, stroke, angina, asthma
and diabetes, self-reported
AC

blood pressure treatment,

smoking, leisure-time
physical activity and
education

Syddall et al. 2003 Hertfordshire Ageing Study, UK 717 67.5 58.6 4 and 10 Handgrip strength Age 52
months
 ACCEPTED MANUSCRIPT

Age, y
Study, year (reference) Study name, country and race Sample size Men, % Follow-up, y Muscle strength test Adjustment for Covariates Number of deaths
(mean or range)

Takata et al. 2007 Study of Fukuoka Prefecture 1282 ≥50 39.7 4 Knee extension and Gender, smoking, BMI, 107
residents, Japan handgrip strength systolic blood pressure,
marital status, levels of total
serum cholesterol or glucose,

PT
or complications from
prevalent diseases

RI
Takata et al. 2012 Japanese 70-year-old community- 600 70.0 51.0 10 Knee extension and Gender, BMI, serum level of 80
dwelling population, Japan handgrip strength total cholesterol, smoking,
and history of cardiovascular
diseases

SC
Taniguchi et al. 2016 Japan 1048 71.6 43.0 4.5 Handgrip strength Gender, age, and baseline 89
year

U
Timpka et al. 2014 Sweden 38 588 19.0 100 42 Knee extension, elbow BMI, smoking, alcohol, 3390
flexion, and handgrip cardiorespiratory fitness,

AN
strength education, coronary heart
disease or stroke diagnosis,
parental socioeconomic status
instead of the subject’s

M
education as follow-up starts
in adolescence

D
van den Beld et al. 2005 The Zoetermeer study, The 403 77.8 100 8 Handgrip strength Age 179
Netherlands

TE
Veronese et al. 2016 Progetto Veneto Anziani Study 2096 75.2 42 4 Knee extension and Age, preserved activities of 327
handgrip strength daily living, cognition,
presence at the baseline of
EP
diabetes, hypertension,
orthostatic hypotension,
cardiovascular diseases,
chronic obstructive
C

pulmonary disease, and

cancer, renal function,
smoking habits, geriatric
AC

nutritional risk index and

number of medications

Willcox et al. 2006 Honolulu Heart Program/Honolulu 5820 54.0 100 43 Handgrip strength Age 3369
Asia Aging Study, USA

BMI, body mass index; MVPA, moderate-to-vigorous physical activity; NA=not available.
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
M
D
TE
EP
C
AC