YCLNU3602_proof ■ 7 September 2018 ■ 1/7

Clinical Nutrition xxx (2018) 1e7

55
Contents lists available at ScienceDirect 56
57
58
Clinical Nutrition 59
60
journal homepage: http://www.elsevier.com/locate/clnu 61
62
63
Meta-analyses 64
65
1 Association between phase angle from bioelectrical impedance 66
2 67
3 analysis and level of physical activity: Systematic review and 68
4
5 Q7 meta-analysis 69
70
6 71
7 Q6 Eduardo Mundstock a, h, Marina Azambuja Amaral b, c, Rafael R. Baptista d, 72
8 Edgar E. Sarria e, Rejane Rosaria Grecco dos Santos a, Adriano Detoni Filho a, 73
9 Carlos Alberto S. Rodrigues a, Gabriele Carra Forte b, Luciano Castro d, 74
10
11
Alexandre Vontobel Padoin b, Ricardo Stein f, Lisiane Marçal Perez a, 75
76
12 Patrícia Klarmann Ziegelmann g, Rita Mattiello a, b, g, * 77
13 a
Programa de Po
s-graduaça ~o em Pediatria e Saúde da Criança, Pontifícia Universidade Cato
lica do Rio Grande do Sul, Porto Alegre, Brazil 78
14 b
Programa de Po
s-graduaça ~o em Medicina e Ci^ encias da Saúde da Pontifícia Universidade Cato
lica do Rio Grande do Sul, Porto Alegre, Brazil 79
15
c
Centro Universita
rio Ritter dos Reis e UniRitter, Porto Alegre, Brazil
d ~o Física e Ci^
lica do Rio Grande do Sul, Porto Alegre, Brazil
80
Faculdade de Educaça encias do Desporto da Pontifícia Universidade Cato
16 e
cia, Curso de Medicina, Universidade de Santa Cruz, Santa Cruz do Sul, Brazil 81
Departamento de Biologia e Farma
17 f
Serviço de Fisiatria e Reabilitaça~o, Hospital de Clínicas de Porto Alegre, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil 82
18 g
Programa de Po
s-graduaça ~o em Epidemiologia, Universidade Federal do Rio Grande do Sul, Porto Alegre, Brazil 83
h
19 Q1 Prefeitura Municipal de Canela/RSeProjeto Esporte e Saúde em Canela (PESC), Brazil
84
20 85
21 86
22 a r t i c l e i n f o s u m m a r y 87
23 88
24 Article history: Objective: To evaluate the relationship between physical activity and phase angle. 89
Received 2 June 2018 Design: Systematic Review and Meta-analysis.
25 90
Accepted 21 August 2018 Data sources: Electronic searches of MEDLINE (via PUBMED), EMBASE, Cochrane Central Register of
26 91
27 Controlled Trials (CENTRAL, The Cochrane Library), SciELO, LILACS, SPORTDiscus, Scopus, and Web of
Keywords: Science from inception to December 10th, 2017.
92
28 Phase angle 93
Eligibility criteria for selecting studies: The PICOS strategy was defined, in which “P” corresponded to
29 Physical activity 94
participants of any age, sex or ethnicity, “I” indicated any type of physical activity program, “C” denoted
30 Bioimpedance
lack of exercise or irregular physical activity, “O” corresponded to the phase angle obtained by bio- 95
Exercise
31 impedance, and “S” indicated longitudinal or cross-sectional studies. 96
Q2 Motor activity
32 Results: In cross-sectional studies the phase angle was higher among the active individuals (MD ¼ 0.70; 97
33 95% CI: 0.48, 0.92, P < 0.001), with low heterogeneity (I2 ¼ 0%; P ¼ 0.619). In longitudinal studies, the 98
34 mean of the difference of phase angles from the baseline was significantly higher for the active group 99
35 than the control group (MD ¼ 0.30; 95% CI: 0.11, 0.49, P ¼ 0.001), with low heterogeneity (I2 ¼ 13%, 100
36 P ¼ 0.331). No evidence of publication bias was found and the overall risk of bias was moderate to high.
101
37 Summary/conclusion: The positive association of physical activity with phase angle reinforces the
102
importance of routinely including exercise in health care. We also identified the need for further studies
38 103
to define with different types, intensities and frequencies of exercises should be conducted in order to
39 104
find the best dose-effect relationship.
40 © 2018 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. 105
41 106
42 107
43 108
44 109
45 110
46 1. Introduction 111
47 112
48 Phase angle (PhA) is considered to be an important parameter, 113
49 * Corresponding author. Pontifícia Universidade Cato
lica do Rio Grande do Sul which is calculated directly from the primary values of the resis- 114
50 (PUCRS), Avenida Ipiranga 6690, Second Floor, Porto Alegre 90610e000, RS, Brazil. tance and reactance of the bio-impedance, in clinical use, as it ap- 115
51 Fax: þ55 51 3320 3312. pears to reflect cellular health, body cell mass, and the integrity of 116
E-mail address: rita.mattiello@pucrs.br (R. Mattiello).
52 117
53 118
https://doi.org/10.1016/j.clnu.2018.08.031
54 0261-5614/© 2018 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved. 119

Please cite this article in press as: Mundstock E, et al., Association between phase angle from bioelectrical impedance analysis and level of
physical activity: Systematic review and meta-analysis, Clinical Nutrition (2018), https://doi.org/10.1016/j.clnu.2018.08.031
 YCLNU3602_proof ■ 7 September 2018 ■ 2/7

2 E. Mundstock et al. / Clinical Nutrition xxx (2018) 1e7

1 phase angle obtained by bio-impedance, and “S” indicated longi- 66

2 Summary box tudinal or cross-sectional studies. 67
3 68
4 2.3. Exclusion criteria 69
5 What is already known 70
6 The following exclusion criteria were used: (A) duplicated 71
7 - Phase Angle reflects cellular health and integrity of the cell publications or studies additional to those already included; (B) 72
8 wall and it has been postulated as a prognostic factor in studies which did not evaluate the association between PA and 73
9 critically ill patients PhA; (C) case studies, case series, experimental models, reviews, 74
10 - Adequate level of physical activity is key to prevent and responses, and editorials; (D) studies without control group; and 75
11 treat chronic diseases (E) studies reporting correlation between physical activity and 76
12 phase angle without classifying the subjects into active and 77
13 inactive. 78
14 What are the new findings 79
15 2.4. Sources of information and search strategy 80
16 - Physical activity has a positive association with phase 81
17 angle. We used the following sources to identify relevant studies from 82
18 inception to December 10th, 2017: MEDLINE (via PUBMED), 83
19 EMBASE, Cochrane Central Register of Controlled Trials (CENTRAL, 84
20 The Cochrane Library), SciELO, LILACS, SPORTDiscus, Scopus, and 85
21 Web of Science. We tried to identify additional studies by scanning 86
the cell wall [1]. Further, ill patients appear to have lower PhA
22 through the reference lists of relevant publications and reviews. 87
values than healthy individuals, which may predict worse health
23 Studies published in any language were included in the analysis. 88
outcomes [2,3] and mortality [4]. Therefore, PhA has been proposed
24 The detailed search strategy can be found online as supple- 89
as a prognostic factor for the diagnosis and monitoring of various
25 mentary material: http://www.crd.york.ac.uk/PROSPEROFILES/ 90
diseases [5e8]. For instance, lower PhA appears to be a prognostic
26 20214_STRATEGY_20151118.pdf. 91
factor predicting mortality in patients with liver cirrhosis [2], un-
27 92
dergoing hemodialysis [9e11], with chronic obstructive pulmonary
28 2.5. Study Selection 93
disease [12], or with cancer [13e15]. It is also associated with
29 94
malnutrition [16e18].
30 To select the studies to be assessed further, two reviewers (EM 95
It is widely acknowledged that physical exercise has positive
31 and MA) independently scanned the titles and abstracts of each 96
effects on the prognosis of all the conditions previously mentioned
32 identified study. The full texts of all potentially relevant articles 97
[19e21]. Maintenance of an adequate level of physical activity (PA)
33 were accessed and investigated. A third reviewer (RM) reviewed 98
is essential in the prevention and treatment of chronic diseases
34 any differences in opinion, to make a final decision. 99
[21]. In addition, engaging in exercise programs improves the
35 100
prognosis of cancers, chronic kidney diseases, and heart conditions
36 2.6. Data extraction 101
[20,22e24]. Evidence suggests that active people, even those with
37 102
chronic disease, have better PhA [25,26], and as a consequence,
38 Data from the selected studies were extracted independently by 103
better health outcomes. However, research data on modifiable
39 two reviewers (EM and MA), using data extraction forms created by 104
factors which may help maintain or increase PhA are scarce.
40 the authors for this review. The following data were extracted: 105
The aim of this study was to evaluate the association between
41 name of the first author, year of publication, study location, pop- 106
physical activity and phase angle, and to conduct a meta-analysis of
42 ulation, type of study, type of exercise assessment (objective or 107
the findings. In order to increase the precision of the results, only
43 subjective), main results, and sample size. 108
studies with a control group, i.e. individuals allocated in a control
44 109
group without exercise or subjects that do not reach the recom-
45 2.7. Type of exercise assessment 110
mended levels of moderate to intense physical activity, were
46 111
included.
47 Physical Activity has four dimensions (Mode or Type, Frequency, 112
48 Duration, and Intensity) and four domains (Occupational, Domestic, 113
49 2. Methods Transportation, and Leisure). These dimensions and domains can be 114
50 objectively verified with wearable technology and monitoring the 115
51 2.1. Protocol and registration vital signs. Physical activity can also be assessed in a subjective 116
52 manner using validated reminders and questionnaires. The choice 117
53 This systematic review and meta-analysis were performed in of the objective or subjective method of PA verification depends on 118
54 accordance with the PRISMA guidelines [27]. The systematic review the level of precision required, on the dimensions and domains to 119
55 protocol was registered with the PROSPERO database: (https://www. be checked, and on the value available for the investment [28]. 120
56 crd.york.ac.uk/prospero/display_record.php?RecordID¼20214). 121
57 2.8. Risk of bias in individual studies 122
58 2.2. Eligibility criteria 123
59 To analyze the risk of bias in clinical trials, we used the Cochrane 124
60 This systematic review and meta-analysis were conducted in Collaboration's tool [29]. We analyzed the degree of randomization, 125
61 order to answer the question: Is physical activity associated with a allocation concealment, blinding of participants, selection bias, 126
62 higher phase angle? The PICOS strategy was defined, in which “P” blinding the researchers analyzing the outcomes, and report of the 127
63 corresponded to participants of any age, sex or ethnicity, “I” indi- selective outcome. 128
64 cated any type of physical activity program, “C” denoted lack of In observational studies, we used an evaluation tool for assess- 129
65 exercise or irregular physical activity, “O” corresponded to the ment of cohort and cross-sectional studies recommended by the 130

Please cite this article in press as: Mundstock E, et al., Association between phase angle from bioelectrical impedance analysis and level of
physical activity: Systematic review and meta-analysis, Clinical Nutrition (2018), https://doi.org/10.1016/j.clnu.2018.08.031
 YCLNU3602_proof ■ 7 September 2018 ■ 3/7

E. Mundstock et al. / Clinical Nutrition xxx (2018) 1e7 3

1 National Institute of Health, U.S. Department of Health and Human 3. Results 66

2 Services [30]. 67
3 The initial search returned 13.546 studies, of which 8.898 were 68
4 2.9. Summary measures and data synthesis duplicate. We screened the remaining 4.648 titles and abstracts, of 69
5 which 4.628 articles were excluded: 4.170 because they did not 70
6 A meta-analysis was conducted separately for cross-sectional evaluate the association between PA and PhA, and 458 because they 71
7 and longitudinal study designs (clinical trials and longitudinal were case studies, cases series, experimental models, reviews, re- 72
8 studies). Mean of differences in the effective size measure was used sponses, and editorials. Twenty articles were selected for full-text 73
9 to estimate the association between physical activity and phase reading, of which 11 were excluded: six due to the lack of a con- 74
10 angle. Mean of differences in the phase angle values was used for trol group and five because they only reported a correlation be- 75
11 cross-sectional studies, while the mean difference of deltas (change tween physical activity and phase angle. Thus, nine articles were 76
12 from baseline) was used for longitudinal studies. In case of studies ultimately included in our systematic review and meta-analyses 77
13 where the standard deviations of deltas were missing, imputation (Fig. 1). 78
14 was performed using a conservative approach (using correlation The general characteristics of the selected studies are summa- 79
15 equal to zero). The Cochran Q test and the I2 statistics were used to rized in Table 1. Taken together, the studies reported data from 575 80
16 assess heterogeneity among studies. Random-effects models on subjects. Five studies were longitudinal: four clinical trials and one 81
17 two separate sub-groups were designed in order to explore the follow-up [26,31e34], and four studies were cross-sectional 82
18 expected heterogeneity among studies. The first sub-group was [25,35e37]. Among the clinical trial studies, three were Random- 83
19 made according to health status (subjects with chronic conditions ized Clinical Trials (RCT) [31e33] and two were Non-Randomized 84
20 or healthy subjects), and the second according to the method used Clinical Trials [26,34]. The clinical trial study by Zanelli et al. was 85
21 to verify PA levels of individuals (objective or subjective measure- designed to evaluate the effect of creatine supplementation on 86
22 ment). The sensitivity was analyzed based on whether exclusion of resistance training practitioners. We used the baseline data for each 87
23 one study at a time significantly modified the heterogeneity or the group for our meta-analysis. The majority of the studies were 88
24 results, and by considering study design in longitudinal studies conducted in different European countries [25,26,35e37]. The au- 89
25 (observational follow-up, randomized clinical trial, or non- thors used different devices to perform the bioimpedance test 90
26 randomized clinical trial). A funnel plot was generated to investi- (Table S2). 91
27 gate the possibility of publication bias. All the analyses were per- The participants (age, athletic condition, and health condition), 92
28 formed using the statistical software ‘R’ version 3.4.0. type of exposure (resistance training, aerobic training, or 93
29 94
30 95
31 96
32 97
33 98
34 99
35 100
36 101
37 102
38 103
39 104
40 105
41 106
42 107
43 108
44 109
45 110
46 111
47 112
48 113
49 114
50 115
51 116
52 117
53 118
54 119
55 120
56 121
57 122
58 123
59 124
60 125
61 126
62 127
63 128
64 129
Fig. 1. Flow diagram of study selection. Q8
65 130

Please cite this article in press as: Mundstock E, et al., Association between phase angle from bioelectrical impedance analysis and level of
physical activity: Systematic review and meta-analysis, Clinical Nutrition (2018), https://doi.org/10.1016/j.clnu.2018.08.031
 YCLNU3602_proof ■ 7 September 2018 ■ 4/7

4 E. Mundstock et al. / Clinical Nutrition xxx (2018) 1e7

1 Table 1 66
2 General characteristics of the selected articles. 67
3 Author, Year Design N Age (years) Health condition Exercise Type Physical activity Country 68
4 assessment 69
5 Ribeiro A S, 2017 [31] Randomized Clinical Trial 50 60 Healthy Resistance training Subjective Brazil 70
6 Souza M F, 2016 [32] Randomized Clinical Trial 41 67.2 ± 4.5 Healthy Resistance training Subjective Brazil 71
7 Martin-Aleman ~ y G, 2016 [33] Randomized Clinical Trial 36 34 (24.5e43) Dialysis Patients Resistance training Subjective Mexico 72
Zanelli J C S, 2015 [34],a Clinical Trial 14 22.6 ± 1.45 Healthy Resistance training Objective Brazil
8 73
Jungblut SA, 2009 [26] Clinical Trial 42 Exercise ¼ 68 COPD Concurrent training Subjective Germany
9 Control ¼ 72 74
10 Meleleo, D, 2017 [37] Follow-up 219 8e11 Healthy Swimming gymnastics Subjective Italy 75
11 Jaramillo G C, 2013 [35] Cross-sectional 78 63 ± 12 Dialysis Patients Walk or Run Objective Spain 76
12 Cupsiti A, 2010 [25] Cross-sectional 50 Exercise ¼ 53.6 ± 11.6 Dialysis Patients Any Objective Italy 77
Control ¼ 63.3 ± 12.5
13 Marra M, 2009 [36] Cross-sectional 45 Exercise ¼ 18.9 ± 1.7 Anorexia nervosa Ballet Subjective Italy
78
14 Control ¼ 19.4 ± 2.4 79
15 80
Legend: N ¼ Number of Subjects; COPD ¼ Chronic Obstructive Pulmonary Disease.
16 a
Clinical trial designed to verify the effects of creatine supplementation on resistance training practitioners; beginners and experienced. For our meta-analysis we used the 81
17 baseline data for each group. 82
18 83
19 84
concurrent training), and methods for assessment of physical ac- among cross-sectional and longitudinal studies (Supplementary
20 85
tivity (subjective and objective) were observed to be heterogenous. Figs. S2 and S3).
21 86
Four studies were conducted on subjects with chronic condi-
22 87
tions [25,26,33,35]. three on patients undergoing dialysis [25,33,35] 3.2. Risk of bias
23 88
and one on patients with chronic obstructive pulmonary disease
24 89
(COPD) [26]. Among the studies performed on healthy populations, The risk of overall bias was moderate to high. Among the clinical
25 90
two were conducted on older women [31,32], one on women with trials, we observed that two studies were at risk of bias in terms of
26 91
anorexia [36], one on resistance training practitioners [34], and one randomization and allocation (Selection Bias). Given the nature of
27 92
on athletic children [37]. The subjects received different types of an exercise intervention, it is not possible to blind participants, but
28 93
training (intervention or exposure). Resistance Training (RT) was the authors of three papers did not declare the blinding with
29 94
more frequently described, with four studies [31e34]. The other respect to the analysis of the outcome (detection bias). None of the
30 95
types of training described were Aerobic Training (AT) [35], con- clinical trials were at risk of bias in relation to incomplete outcome
31 96
current training [26], swimming and gymnastics [37], ballet data (attrition bias) or selective reporting (reporting bias)
32 97
dancing [36], and any form of physical activity [25]. As for the tools (Supplementary Fig. S4).
33 98
used to assess PA in the five observational studies, two studies used Only one of the observational studies had an overall good rating
34 99
objective measurements: one used accelerometers [25], and one in terms of quality, while two were rated as fair and one as poor
35 100
used pedometers [35], while two used subjective tools [36,37]. (Supplementary Table 1).
36 101
Regarding the interventions performed in the five clinical trials, the
37 102
three RCT used resistance training [31e33], one clinical trial used
38 4. Discussion 103
resistance training [34], and one used concurrent training [26].
39 104
40 Our systematic review found evidence that physical activity had 105
41 3.1. Meta-analysis a positive effect on phase angle. These results were corroborated by 106
42 the meta-analysis of data from clinical trials, showing evidence of a 107
43 In the cross-sectional studies, the active subjects presented a causal relationship, wherein differences in PhA from the baseline 108
44 higher mean phase angle than controls (MD ¼ 0.70; 95% CI: 0.48, were significantly higher for the active group than the control 109
45 0.92, P < 0.001, Fig. 2), with low heterogeneity (I2 ¼ 0%; P ¼ 0.619). It group. The magnitude of the increase in phase angle was not sta- 110
46 was also observed that the differences between health status and tistically different between healthy individuals or those with 111
47 method for assessment of physical activity were not significant chronic disease, either by type of study or by type of physical ac- 112
48 (P ¼ 0.332; P < 0.253, Fig. 2). The longitudinal studies (clinical trials tivity evaluation, which shows great consistency of results. 113
49 or follow-up) indicated that the mean differences of the phase Different mechanisms appear to be involved in the process by 114
50 angles from baseline were significantly higher for the active group, which PA causes improvement in PhA values, resulting in better 115
51 when compared with the control group (MD ¼ 0.37; 95% CI: 0.13, integrity and functionality of the cell membrane, changes in 116
52 0.61, P ¼ 0.002, Fig. 2), with low heterogeneity (I2 ¼ 13%, P ¼ 0.331). intracellular composition, and enhanced tissue capacity [1,38]. 117
53 The differences between health status were not significant Although exercise initially appears to negatively affect the cell 118
54 (P ¼ 0.7821, Fig. 2). All studies had subjective measures of the level membrane [39], this is followed by a super compensation in which 119
55 of exercise. There were no significant differences between clinical the acute harmful effect of exercise on the individual is compen- 120
56 trials and follow-up studies (P ¼ 0.502, Fig. 2). A sensitivity analysis sated by an improvement of the structure [40], i.e. comply with the 121
57 was conducted after excluding one study by Martin-Aleman ~ y et al., principle of the general adaptation syndrome [41]. 122
58 which had a high disparity in results [33] (Supplementary Fig. S1). The phase angle is inversely proportional to resistance; resis- 123
59 Although this did not change the overall direction of the effect tance, in turn, depends on both intra and extracellular water [42]. 124
60 (MD ¼ 0.39; 95% CI: 0.19, 0.58, P < 0.001), heterogeneity was Physical training, especially when it causes an increase in muscle 125
61 observed to decrease (I2 ¼ 0%). In a more detailed sensitivity mass, can lead to increase the intracellular water [43], which re- 126
62 analysis in which one study was excluded at a time, the mean duces the resistance and consequently leads to an increase of the 127
63 difference in both cross-sectional and longitudinal studies did not phase angle. The reactance is directly proportional to the phase 128
64 change with the exclusion of any of the studies. Analysis of the angle and depends on the integrity of the cell membrane [42]. Well- 129
65 Funnel Plot did not show a tendency for publication bias, both conducted exercise may be a factor in enhancing the integrity of 130

Please cite this article in press as: Mundstock E, et al., Association between phase angle from bioelectrical impedance analysis and level of
physical activity: Systematic review and meta-analysis, Clinical Nutrition (2018), https://doi.org/10.1016/j.clnu.2018.08.031
 YCLNU3602_proof ■ 7 September 2018 ■ 5/7
E. Mundstock et al. / Clinical Nutrition xxx (2018) 1e7 5
1 66
2 67
3 68
4 69
5 70
6 71
7 72
8 73
9 74
10 75
11 76
12 77
13 78
14 79
15 80
16 81
17 82
18 83
19 84
20 85
21 86
22 87
23 88
24 89
25 90
26 91
27 92
28 93
29 94
30 95

Fig. 2. Meta-analyzes.
31 96
32 97
33 98
34 99
35 100
36 101
37 102
38 103
39 104
40 105
41 106
42 107
43 108
44 109
45 110
46 111
47 112
48 113
49 114
50 115
51 116
52 117
53 118
54 119
55 120
56 121
57 122
58 123
59 124
60 125
61 126
62 127
63 128
64 129
65 130
Please cite this article in press as: Mundstock E, et al., Association between phase angle from bioelectrical impedance analysis and level of
physical activity: Systematic review and meta-analysis, Clinical Nutrition (2018), https://doi.org/10.1016/j.clnu.2018.08.031
 YCLNU3602_proof ■ 7 September 2018 ■ 6/7

6 E. Mundstock et al. / Clinical Nutrition xxx (2018) 1e7

1 this membrane by the mechanism of overcompensation described However, when analyzing the effect size between the studies, there 66
2 above [41]. Another factor that increases the reactance is the total was no clinically significant difference between the studies with 67
3 cellular mass [44]. Training can lead to the increase of the total better or worse quality, which shows that this was not a deter- 68
4 cellular mass [45], which leads to the increase of the reactance and mining factor for the observed results. 69
5 the consequent increase of the phase angle. Measurement of the 70
6 phase angle can, therefore, be an indicator of the effects of physical 4.1. Strengths and limitations of the systematic review 71
7 exercise on the health of the cell and, consequently, on the health of 72
8 the individual. There are also authors who argue that the phase The major strength of our systematic review is the inclusion of 73
9 angle can be used, in clinical practice, as an indicator of the level of all available studies, by searching the “gray” literature as well as all 74
10 physical activity of the individual [46]. major databases. We also did not limit the search by publication 75
11 The type of exercise most studied among the included articles period or by language, furthermore we do not set limits in terms of 76
12 was resistance training. This type of training appears, alone or in age, type of diet and type or level of exercise performed. One of the 77
13 conjunction with other training modalities, in five of the nine ar- problems that this could cause would be a great statistical het- 78
14 ticles. In addition, this training modality appears in all clinical trials. erogeneity among the studies, which was not verified in the meta- 79
15 Resistance training was the only type o exercise in which we found analyzes. 80
16 a study that attempted to identify the effect of changes in the A limitation of our study is the low methodological quality of the 81
17 volume intensity ratio. This study [31] identified that this training studies, as shown in the analysis of risk of bias. This limitation 82
18 modality, when performed in elderly women with different seems to be of little importance, as there is a great consistency in 83
19 methods has positive effects on the phase angle. However, there are the results between the studies with reference to the positive effect 84
20 other methods and different relationships between volume, in- of PA on the PhA. 85
21 tensity and frequency in training against resistance, and especially 86
22 in other types of exercise that must be tested to find the best dose- 5. Conclusions and future directions 87
23 benefit relationship in relation to phase angle. Furthermore, we did 88
24 not find studies investigating whether the combination of more Our study found that physical activity has a positive effect on 89
25 than one type of exercise, the concurrent training, is superior to the phase angle, which is a prognostic marker of cell health. That is one 90
26 isolated training. Therefore, the present study supports the hy- more reason to support the inclusion of physical activity counseling 91
27 pothesis that exercise is positively associated with phase angle and in the clinical routine, treatment, and prevention of various chronic 92
28 therefore leads to enhanced cell membrane integrity and improved diseases. Also, our results suggest that the clinicians may include 93
29 cell health. However, the relationships between domains and di- phase angle in the assessment of physical performance, since its 94
30 mensions and, especially, the best dose-benefit relationship are evaluation, assessed by the BIA method is reliable, easy to measure, 95
31 open questions. and repeatable. 96
32 Although the studies included different populations, types of These results are important for clinical, research, and routine 97
33 interventions, did not specify the recommended criteria for per- scenarios, aiming for the better health and well-being of both 98
34 forming bioimpedance, items that affect the results, the heteroge- healthy individuals and subjects with chronic diseases. Still, addi- 99
35 neity of the analyzes was low and the direction of the effect went to tional studies comparing different types, intensities, and fre- 100
36 the same side. The only study that the direction differed a little quencies of exercises should be conducted in order to find the best 101
37 from the others was that of Martin-Aleman ~ y [33], the exclusion of dose-effect relationship. Q4 102
38 this study increases the effect of the results. These findings rein- 103
39 force that in spite of all possible differences that could affect the Conflict of interest Q9 104
40 results, the effect of the exercise continued to indicate the impor- 105
41 tance of this and suggests that in studies that are more homoge- The authors declare no conflict of interest. 106
42 neous and in ideal scenarios to perform bioimpedance, this effect 107
43 may be even greater. Appendix A. Supplementary data 108
44 Evidence demonstrates that larger PhA values are associated 109
45 with lower risk of hospitalization, mortality, and cardiovascular Supplementary data related to this article can be found at 110
46 events [4,47] in patients with chronic conditions [11,13e15,48]. This https://doi.org/10.1016/j.clnu.2018.08.031. 111
47 substantiates the importance of including PA routines with the aim 112
48 of improving outcomes in patients with chronic diseases. In our References 113
49 systematic review, all four studies in patients with chronic condi- 114
50 tions [25,26,33,35] reported a significant positive association be- [1] Norman K, Stobaus N, Pirlich M, Bosy-Westphal A. Bioelectrical phase angle 115
and impedance vector analysis–clinical relevance and applicability of
51 tween PhA and the level of PA. As the phase angle has been impedance parameters. Clin Nutr 2012;31:854e61. https://doi.org/10.1016/
116
52 postulated as an indicator of cellular health, cell membrane integ- j.clnu.2012.05.008. S0261-5614(12)00108-2 [pii]. 117
53 rity [49] and total cell mass [1], these results indicate that, even in [2] Selberg O, Selberg D. Norms and correlates of bioimpedance phase angle in 118
healthy human subjects, hospitalized patients, and patients with liver
54 patients with chronic diseases, exercise improves these indicators. cirrhosis. Eur J Appl Physiol 2002;86:509e16. https://doi.org/10.1007/s00421-
119
55 The quality of the studies in general was reasonable, but some 001-0570-4. 120
56 items were poorly reported, especially among clinical trials. An [3] Ringaitiene D, Gineityte D, Vicka V, Zvirblis T, Norkiene I, Sipylaite J, et al. 121
Malnutrition assessed by phase angle determines outcomes in low-risk car-
57 unreported item was the procedure to blind the study. The partic- 122
diac surgery patients. Clin Nutr 2016;35:1328e32. https://doi.org/10.1016/
58 ipant in a study with exercise will always know if he is in the j.clnu.2016.02.010. 123
59 control group or not. However, the personnel that analyzes the [4] Garlini LM, Alves FD, Ceretta LB, Perry IS, Souza GC, Clausell NO. Phase angle 124
60 and mortality: a systematic review. Eur J Clin Nutr 2018. https://doi.org/ 125
phase angle can be blinded, which has not happened in some
10.1038/s41430-018-0159-1.
61 studies. Another source of risk of bias is absence of randomization [5] Colin-Ramirez E, Castillo-Martinez L, Orea-Tejeda A, Vazquez-Duran M, 126
62 in some studies, which may generate a selection bias, since in- Rodriguez AE, Keirns-Davis C. Bioelectrical impedance phase angle as a 127
63 dividuals with a lower phase angle could be placed in the exercise prognostic marker in chronic heart failure. Nutrition 2012;28:901e5. https:// 128
doi.org/10.1016/j.nut.2011.11.033. S0899-9007(11)00442-4 [pii].
64 group, which would predispose this group to greater increases in [6] Visser M, van Venrooij LM, Wanders DC, de Vos R, Wisselink W, van 129
65 this variable due to the phenomenon of return to the mean [50]. Leeuwen PA, et al. The bioelectrical impedance phase angle as an indicator of 130

Please cite this article in press as: Mundstock E, et al., Association between phase angle from bioelectrical impedance analysis and level of
physical activity: Systematic review and meta-analysis, Clinical Nutrition (2018), https://doi.org/10.1016/j.clnu.2018.08.031
 YCLNU3602_proof ■ 7 September 2018 ■ 7/7

E. Mundstock et al. / Clinical Nutrition xxx (2018) 1e7 7

1 undernutrition and adverse clinical outcome in cardiac surgical patients. Clin A scientific statement from the American Heart Association. Circulation 66
Nutr 2012;31:981e6. https://doi.org/10.1016/j.clnu.2012.05.002. S0261- 2013;128:2259e79.
2 67
5614(12)00102-1 [pii]. [29] Higgins JPT, Green S. Cochrane handbook for systematic reviews of in-
3 [7] Grundmann O, Yoon SL, Williams JJ. The value of bioelectrical impedance terventions, vol. 5. Wiley Online Library; 2008. 68
4 analysis and phase angle in the evaluation of malnutrition and quality of life in [30] National Institution of Health, U.S Department of Health and Human Services. 69
5 cancer patients-a comprehensive review. Eur J Clin Nutr 2015;69:1290e7. Quality assessment tool for observational cohort and cross-sectional studies. 70
https://doi.org/10.1038/ejcn.2015.126. ejcn2015126 [pii]. 2014. https://www.nhlbi.nih.gov/health-pro/guidelines/in-develop/cardiov
6 [8] S. T, K. A, K. K, T. H, K. I, M. T, et al. A low phase angle measured with ascular-risk-reduction/tools/cohort. [Accessed 1 December 2017]. Q5
71
7 bioelectrical impedance analysis is associated with osteoporosis and is a risk [31] Ribeiro AS, Schoenfeld BJ, Souza MF, Tomeleri CM, Silva AM, Teixeira DC, et al. 72
8 factor for osteoporosis in community-dwelling people: the Yakumo study. Resistance training prescription with different load-management methods 73
Arch Osteoporos 2018;13. https://doi.org/10.1007/s11657-018-0450-8. improves phase angle in older women. Eur J Sport Sci 2017;17:913e21.
9 [9] Maggiore Q, Nigrelli S, Ciccarelli C, Grimaldi C, Rossi GA, Michelassi C. Nutri- https://doi.org/10.1080/17461391.2017.1310932.
74
10 tional and prognostic correlates of bioimpedance indexes in hemodialysis [32] Souza MF, Tomeleri CM, Ribeiro AS, Schoenfeld BJ, Silva AM, Sardinha LB, et al. 75
11 patients. Kidney Int 1996;50:2103e8. Effect of resistance training on phase angle in older women: a randomized 76
[10] Fein PA, Gundumalla G, Jorden A, Matza B, Chattopadhyay J, Avram MM. controlled trial. Scand J Med Sci Sports 2017;27:1308e16. https://doi.org/
12 Usefulness of bioelectrical impedance analysis in monitoring nutrition status 10.1111/sms.12745.
77
13 and survival of peritoneal dialysis patients. Adv Perit Dial 2002;18:195e9. [33] Martin-Aleman ~ y G, Valdez-Ortiz R, Olvera-Soto G, Gomez-Guerrero I, Aguire- 78
14 [11] Beberashvili I, Azar A, Sinuani I, Shapiro G, Feldman L, Stav K, et al. Bio- Esquivel G, Cantu-Quintanilla G, et al. The effects of resistance exercise and 79
impedance phase angle predicts muscle function, quality of life and clinical oral nutritional supplementation during hemodialysis on indicators of nutri-
15 80
outcome in maintenance hemodialysis patients. Eur J Clin Nutr 2014;68: tional status and quality of life. Nephrol Dial Transplant 2016;31:1712e20.
16 683e9. https://doi.org/10.1038/ejcn.2014.67. ejcn201467 [pii]. https://doi.org/10.1093/ndt/gfw297. 81
17 [12] Maddocks M, Kon SS, Jones SE, Canavan JL, Nolan CM, Higginson IJ, et al. [34] Zanelli JCS, Cordeiro BA, Beserra BTS, Trindade EBS de M. Creatine and 82
Bioelectrical impedance phase angle relates to function, disease severity and resistance training: effect on hydration and lean body mass. Rev Bras Med Do
18 83
prognosis in stable chronic obstructive pulmonary disease. Clin Nutr 2015;34: Esporte 2015;21:27e31. https://doi.org/10.1590/1517-86922015210101932.
19 1245e50. https://doi.org/10.1016/j.clnu.2014.12.020. S0261-5614(15)00003- [35] Jaramillo GC, Gallar P, Di Gioia C, Rodriguez I, Ortega O, Herrero JC, et al. Basal 84
20 5 [pii]. physical activity in hemodialysis patients. Correlation with biochemical pa- 85
21 [13] Gupta D, Lis CG, Dahlk SL, Vashi PG, Grutsch JF, Lammersfeld CA. Bioelectrical rameters and with body composition. Nephrol Dial Transplant 2013;28:485. 86
impedance phase angle as a prognostic indicator in advanced pancreatic [36] Marra M, Caldara A, Montagnese C, De Filippo E, Pasanisi F, Contaldo F, et al.
22 cancer. Br J Nutr 2004;92:957e62. doi:S0007114504002594 [pii]. Bioelectrical impedance phase angle in constitutionally lean females, ballet 87
23 [14] Gupta D, Lammersfeld CA, Vashi PG, King J, Dahlk SL, Grutsch JF, et al. dancers and patients with anorexia nervosa. Eur J Clin Nutr 2009;63:905e8. 88
24 Bioelectrical impedance phase angle as a prognostic indicator in breast cancer. [37] Meleleo D, Bartolomeo N, Cassano L, Nitti A, Susca G, Mastrototaro G, et al. 89
BMC Cancer 2008;8:249. https://doi.org/10.1186/1471-2407-8-249. 1471- Evaluation of body composition with bioimpedance. A comparison between
25 2407-8-249 [pii]. athletic and non-athletic children. Eur J Sport Sci 2017;17:710e9. https://
90
26 [15] Paiva SI, Borges LR, Halpern-Silveira D, Assuncao MC, Barros AJ, Gonzalez MC. doi.org/10.1080/17461391.2017.1291750. 91
27 Standardized phase angle from bioelectrical impedance analysis as prognostic [38] Barbosa-Silva MC, Barros AJ. Bioelectrical impedance analysis in clinical 92
factor for survival in patients with cancer. Supp Care Cancer 2009;19:187e92. practice: a new perspective on its use beyond body composition equations.
28 https://doi.org/10.1007/s00520-009-0798-9. Curr Opin Clin Nutr Metab Care 2005;8:311e7. doi:00075197-200505000-
93
29 [16] Slee A, Birc D, Stokoe D. Bioelectrical impedance vector analysis, phase-angle 00012 [pii]. 94
30 assessment and relationship with malnutrition risk in a cohort of frail older [39] Perez AC, de Oliveira ACC, Estevez E, Molina AJ, Prieto JG, Alvarez AI. Mito- 95
hospital patients in the United Kingdom. Nutrition 2014;31:132e7. https:// chondrial, sarcoplasmic membrane integrity and protein degradation in heart
31 96
doi.org/10.1016/j.nut.2014.06.002. S0899-9007(14)00284-6 [pii]. and skeletal muscle in exercised rats. Comp Biochem Physiol Part C Toxicol
32 [17] Kyle UG, Soundar EP, Genton L, Pichard C. Can phase angle determined by Pharmacol 2003;134:199e206. 97
33 bioelectrical impedance analysis assess nutritional risk? A comparison be- [40] Issurin VB. Generalized training effects induced by athletic preparation. A 98
tween healthy and hospitalized subjects. Clin Nutr 2012;31:875e81. https:// review. J Sports Med Phys Fit 2009;49:333e45.
34 99
doi.org/10.1016/j.clnu.2012.04.002. [41] Cunanan AJ, DeWeese BH, Wagle JP, Carroll KM, Sausaman R, Hornsby WG,
35 [18] Kyle UG, Genton L, Pichard C. Low phase angle determined by bioelectrical et al. The general adaptation syndrome: a foundation for the concept of 100
36 impedance analysis is associated with malnutrition and nutritional risk at periodization. Sports Med 2018;48:787e97. https://doi.org/10.1007/s40279- 101
37 hospital admission. Clin Nutr 2013;32:294e9. https://doi.org/10.1016/ 017-0855-3. 102
j.clnu.2012.08.001. [42] Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M, Gomez JM, et al.
38 [19] Speck RM, Courneya KS, Masse LC, Duval S, Schmitz KH. An update of Bioelectrical impedance analysis–part I: review of principles and methods. 103
39 controlled physical activity trials in cancer survivors: a systematic review and Clin Nutr 2004;23:1226e43. https://doi.org/10.1016/j.clnu.2004.06.004. 104
40 meta-analysis. J Cancer Surviv 2010;4:87e100. https://doi.org/10.1007/ S0261561404000937 [pii]. 105
s11764-009-0110-5. [43] Ribeiro AS, Avelar A, Schoenfeld BJ, Ritti Dias RM, Altimari LR, Cyrino ES.
41 [20] Barcellos FC, Santos IS, Umpierre D, Bohlke M, Hallal PC. Effects of exercise in Resistance training promotes increase in intracellular hydration in men and
106
42 the whole spectrum of chronic kidney disease: a systematic review. Clin women. Eur J Sport Sci 2014;14:578e85. https://doi.org/10.1080/ 107
43 Kidney J 2015;8:753e65. https://doi.org/10.1093/ckj/sfv099 sfv099. 17461391.2014.880192. 108
[21] Pedersen BK, Saltin B. Exercise as medicine - evidence for prescribing exercise [44] Goswami PN, Munna K. Bioelectrical impedance analysis: phase angle - an
44 as therapy in 26 different chronic diseases. Scand J Med Sci Sport independent predictive health marker and its clinical applications. Water
109
45 2015;25(Suppl 3):1e72. https://doi.org/10.1111/sms.12581. 2007;15(2007). https://doi.org/10.1007/978-3-540-68017-8_82. 110
46 [22] Bourke L, Smith D, Steed L, Hooper R, Carter A, Catto J, et al. Exercise for men [45] Pigłowska M, Kostka T, Drygas W, Jegier A, Leszczyn
ska J, Bill-Bielecka M, et al. 111
with prostate cancer: a systematic review and meta-analysis. Eur Urol Body composition, nutritional status, and endothelial function in physically
47 112
2016;69:693e703. https://doi.org/10.1016/j.eururo.2015.10.047. active men without metabolic syndrome e a 25 year cohort study. Lipids
48 [23] Lavie CJ, Lee DC, Sui X, Arena R, O'Keefe JH, Church TS, et al. Effects of running Health Dis 2016;15:1e8. https://doi.org/10.1186/s12944-016-0249-9. 113
49 on chronic diseases and cardiovascular and all-cause mortality. Mayo Clin [46] Mann S, Wade M, Fisher J, Giessing J, Gentil P, Steele J. Phase angle as an 114
Proc 2015;90:1541e52. https://doi.org/10.1016/j.mayocp.2015.08.001. indicator of health and fitness in patients entering an exercise referral
50 115
S0025-6196(15)00621-7 [pii]. scheme. J Am Med Dir Assoc 2018:7e8. https://doi.org/10.1016/
51 [24] Pothirat C, Chaiwong W, Phetsuk N, Liwsrisakun C, Bumroongkit C, j.jamda.2018.06.005. 116
52 Deesomchok A, et al. Long-term efficacy of intensive cycle ergometer exercise [47] de Araujo Antunes A, Vannini FD, de Arruda Silveira LV, Barretti P, Martin LC, 117
53 training program for advanced COPD patients. Int J Chronic Obstr Pulm Dis JCT Caramori. Associations between bioelectrical impedance parameters and 118
2015;10:133e44. https://doi.org/10.2147/COPD.S73398. cardiovascular events in chronic dialysis patients. Int Urol Nephrol 2013;45:
54 [25] Cupisti A, Capitanini A, Betti G, D'Alessandro C, Barsotti G. Assessment of 1397e403. 119
55 habitual physical activity and energy expenditure in dialysis patients and [48] do Amaral Paes TC, de Oliveira KCC, de Carvalho Padilha P, Peres WAF. Phase 120
56 relationships to nutritional parameters. Clin Nephrol 2010;75:218e25. doi: angle assessment in critically ill cancer patients: relationship with the 121
8445 [pii]. nutritional status, prognostic factors and death. J Crit Care 2018;44:430e5.
57 [26] Jungblut SA, Frickmann H, Zimmermann B, Muller U, Bargon J. The effects of https://doi.org/10.1016/j.jcrc.2018.01.006.
122
58 physical training on the body composition of patients with COPD. Pneumo- [49] Kumar S, Dutt A, Hemraj S, Bhat S, Manipadybhima B. Phase angle measure- 123
59 logie 2009;63:374e9. https://doi.org/10.1055/s-0029-1214713. ment in healthy human subjects through bio-impedance analysis. Iran J Basic 124
[27] Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for sys- Med Sci 2012;15:1180e4.
60 tematic reviews and meta-analyses: the PRISMA statement. BMJ 2009;339. [50] Streiner DL. Statistics commentary series: commentary #16-regression to-
125
61 https://doi.org/10.1136/bmj.b2535. b2535eb2535. ward the mean. J Clin Psychopharmacol 2016;36:416e8. https://doi.org/ 126
62 [28] Strath SJ, Kaminsky LA, Ainsworth BE, Ekelund U, Freedson PS, Gary RA, et al. 10.1097/JCP.0000000000000551. 127
Guide to the assessment of physical activity: clinical and research applications
63 128
64 129
65 130

Please cite this article in press as: Mundstock E, et al., Association between phase angle from bioelectrical impedance analysis and level of
physical activity: Systematic review and meta-analysis, Clinical Nutrition (2018), https://doi.org/10.1016/j.clnu.2018.08.031