Spontaneous Cardiac-Locomotor Coupling in Healthy Individuals

During Daily Activities

Aurora Rosato

1 a

, Matilda Larsson

1 b

, Eric Rullman

2 c

and Seraina A. Dual

1 d

Department of Biomedical Engineering, KTH Royal Institute of Technology, Stockholm, Sweden

Division of Clinical Physiology, Department of Laboratory Medicine, Karolinska Institutet,

Karolinska University Hospital Huddinge, Stockholm, Sweden

Keywords:

Cardiac-Locomotor Coupling, Wearable Sensor, Exercise, Lifelogging.

Abstract:

During exercise, the locomotor and the cardiovascular system work in synergy to control the blood ﬂow

through the body. In particular, the muscle contraction generates rhythmic raising and lowering of intramus-

cular pressure, which in synergy supports cardiovascular function. This study aims to analyze spontaneous

cardiac-locomotor coupling (CLC) events during daily activities using weareable sensors. We analyze the data

set PMData, containing recordings from sixteen healthy subjects during ﬁve months. The data were acquired

with a smartwatch and consist of step rate (SR), heart rate (HR) and daily surveys reporting the training ses-

sions. Coupling is deﬁned as being present when SR and HR are within 1% of each other (strong coupling)

and within the 10% of each other (weak coupling). The results show that every subject presents occurrences

of CLC while performing normal daily activities. In particular, strong coupling occurs more likely for longer

activities (111 ± 34 min), at moderate intensity (100

steps

min

< SR > 130

steps

min

). The presence of CLC during

daily activities rises the question whether there is a physiological mechanism controlling this phenomenon,

that should be investigated in future.

1 INTRODUCTION

The human body is made of a set of systems that work

in synergy to achieve maximum efﬁciency, within cer-

tain metabolic and bio-mechanical constraints. The

cardiovascular, the respiratory and the locomotor sys-

tem affect blood ﬂow.

In fact, during locomotion, blood ﬂow through

the body is inﬂuenced by two opposing pumps. The

pumping of the heart delivers the blood to the whole

body and the skeletal muscle pump, through periodic

increases of intramuscular pressure and venous return

(Novak et al., 2007), pumps it back to the heart. If

these two pumps become entrained, with equal con-

traction rates, the cardiac-locomotor coupling (CLC)

phenomenon occurs (Niizeki and Saitoh, 2014).

Prior studies of CLC investigate the interaction

between the cardiovascular and the locomotor sys-

tem during rhythmic exercise in the laboratory setting

https://orcid.org/0000-0002-8768-2619

https://orcid.org/0000-0002-5795-9867

https://orcid.org/0000-0003-2854-7262

https://orcid.org/0000-0001-6867-8270

(Kirby et al., 1989; Hausdorff et al., 1992; Constan-

tini et al., 2018; Takeuchi et al., 2014). In the past, re-

searchers used different signal processing techniques

to identify the CLC during exercise. One method con-

sists in processing the electrocardiogram (ECG) and

the acceleration, identifying the rates and calculating

the ratio between the heart rate (HR) and the step rate

(SR) (Kirby et al., 1989). Another technique consists

in analyzing the frequency spectrum and the coher-

ence of the acquired signals (Hausdorff et al., 1992;

Niizeki et al., 1993). It has been demonstrated that

the HR increases when the muscle contraction is syn-

chronized with the systolic phase of the cardiac cycle

(Niizeki and Miyamoto, 1999). In contrast, if oppor-

tunely synchronized with diastole (see Figure 1), it

can reduce the HR and the ventilation, indicating im-

proved cardiac efﬁciency (Constantini et al., 2018).

All prior studies on CLC take place in controlled,

monitored environments and study short-term activ-

ities such as walking and running on a treadmill or

cycling (Nomura et al., 2003). However, the physi-

ological occurrence of CLC remains elusive. In par-

ticular, we do not know if humans in their daily lives

experience CLC, a necessary condition to allow for

170

Rosato, A., Larsson, M., Rullman, E. and Dual, S.

Spontaneous Cardiac-Locomotor Coupling in Healthy Individuals During Daily Activities.

DOI: 10.5220/0011632700003414

In Proceedings of the 16th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2023) - Volume 4: BIOSIGNALS, pages 170-177

ISBN: 978-989-758-631-6; ISSN: 2184-4305

 2023 by SCITEPRESS – Science and Technology Publications, Lda. Under CC license (CC BY-NC-ND 4.0)

diastolic synchronization. Our hypothesis is that the

CLC is a common occurrence across individuals, not

only in laboratory settings but also during activities of

daily life.

The use of smartphones and wearables has be-

come common practice to record signals about phys-

ical activities and health status, called lifelogging

(Karami et al., 2021). Exploiting this trend, this

study aims to use lifelog data to analyze the extent

of CLC occurrences during non-monitored daily ac-

tivities, collected retrospectively from a wrist-worn

wearable device. Moreover, we want to investigate

how occurrence of CLC depends on the intensity, the

type and the duration of the activity and subject-age.

Cardiac cycle

ECG

Figure 1: Timing of cardiac cycle and gait cycle during di-

astolic cardiac-locomotor coupling.

2 METHODS

2.1 Study Population

We used a retrospectively acquired dataset called PM-

Data (Thambawita et al., 2020). The dataset contains

data collected from 16 healthy subjects (13 men and

3 women), aged between 23 and 60 years (34.85 ±

11.67 years) (Table 1).

The participants were grouped into elderly and

young to assess inter-class differences using a com-

mon threshold of 40 years of age. The elderly group

presents 5 subjects (49.80 ± 7.22 years) and the young

group 11 subjects (28.09 ± 4.64 years).

2.2 Study Protocol and Recorded

Variables

The participants wore a Fitbit Versa 2 (Fitbit.Inc, San

Francisco) for a period of 5 months, from November

2019 to the end of March 2020. They were encour-

aged to wear the smartwatch as much as possible. The

study organizers did not impose any restrictions or re-

quirements on the type or duration of exercise (Tham-

bawita et al., 2020).

The Fitbit acquired SR per minute and HR per 5

seconds. Each entry has a timestamp that allows to

synchronize samples from different ﬁles. Moreover,

the participants were encouraged to use PM Reporter

Pro smartphone application (Forzasys AS c/o Simula

Research Laboratory, Oslo) to collect subjective as-

sessments of training load, reported after every train-

ing session. The subjective assessments of training

load are collected in CSV-ﬁles named Session Rating

of Perceived Exertion (SRPE). Each SRPE ﬁle con-

tains the type of activity performed, the training ses-

sion’s end time, the duration and the rate of perceived

exertion (RPE), used to assess the internal training

load.

2.3 Data Analysis

The ﬁles were processed in MATLAB R2022a (The

MathWorks, Natick, MA) for all analyses. We used

the Statistic and Machine Learning Toolbox.

Since the HR and the SR were not acquired with the

same sampling rate, we ﬁrst synchronized both sig-

nals using the timestamp information. The HR was

averaged per minute, in order to match the time res-

olution of the SR and to account for possible ﬂuctua-

tion in HR due to potential ﬂuctuations in device ac-

curacy.

The data were ﬁltered to instances of physi-

cal activity with the goal to delete instances of

sleep and rest. We used the metabolic equivalent

of task (METs) to evaluate physical activity inten-

sity. One metabolic equivalent represents the oxy-

gen consumption while sitting at rest (1 MET= 3.5

ml O

/kg/min)(Jett

e et al., 1990). We used a MET

level of 2 and a threshold of 60

steps

min

as indicative of

instances of physical activity, as this SR was previ-

ously identiﬁed as slow walking in a study population

of adults older than 20 years of age. (Tudor-Locke

et al., 2011). All data below 60

steps

min

were excluded

from the analysis.

2.3.1 Cardiac-Locomotor Coupling

To ﬁnd evidence of CLC, the ratio between SR and

HR, deﬁned as:

R =

SR(

steps

min

)

HR(

beats

min

)

(1)

was computed for each entry. Coupling is deﬁned as a

deviation between SR and HR of < 1% (Kirby et al.,

1989). We group the data into uncoupling (deviation

> 10%), weak coupling (deviation < 10%) and strong

Spontaneous Cardiac-Locomotor Coupling in Healthy Individuals During Daily Activities

171

coupling (deviation < 1%) for further considerations.

We introduced a new parameter to evaluate the devia-

tion from the ideal ratio 1:1. We called this parameter

coupling parameter and we deﬁned it as follow.

C = |

SR(

steps

min

)

HR(

beats

min

)

− 1| (2)

We used violin plot of the SR and the HR to look

at their distribution. We decided to group the subjects

in three groups, according to the number of step rates

at which strong coupling occurred. In group 1, strong

coupling occurred at one speciﬁc step rate, in group 2

at two step rates and in group 3 at three step rates.

2.3.2 Timing, Duration and Intensity

We analyzed the timing, the duration and the intensity

of physical activities. For these calculations, we only

used the dataset with SR and HR, which contains no

information reported by the participant about which

activity was performed.

We deﬁned an activity as consecutive observations

with a difference between them smaller than one hour.

In this way, we were able to investigate the training

habits and to calculate:

• the mean duration of activities

• the time of the day at which the activities were

conducted

Furthermore, we stratiﬁed the data into three

groups according to their heuristic cadence thresh-

olds of 100

steps

min

and 130

steps

min

which are associated to

moderate and vigorous intensity, respectively (Tudor-

Locke et al., 2019; Tudor-Locke et al., 2020; Tudor-

Locke et al., 2021). Hence, we divided the intensity of

any activity in three different groups: light, moderate

and vigorous activity.

2.3.3 Type of Training

We combined the information from the Fitbit and the

one reported in the SRPE, described in the section 2.2.

The reports about the training load contain four vari-

ables: end time of the training, type of training, dura-

tion and RPE. For each entry, the mean coupling pa-

rameter was calculated and the data were grouped ac-

cording to two activities (running and strength train-

ing). Moreover, we computed the percentage of cou-

pling for each training.

2.4 Statistical Analysis

Firstly, we plotted the distribution of the ratio to as-

sess if coupling (1:1) was more likely then any other

ratio. Then, a Chi-squared test was used to determine

if the distribution of the ratio, coupling parameter, HR

and SR is normal. Since these data were not nor-

mally distributed, we used nonparametric statistical

tests to assess differences. Median and standard de-

viation were computed for each parameter of interest.

For the amount of coupling, the timing and the per-

centage of occurrences, mean and standard deviation

were calculated instead.

We chose the Kruskal-Wallis test to test statisti-

cally signiﬁcant differences in grouping the subjects,

according to their preferred SR, as explained in sec-

tion 2.3.1. A Friedman test was instead used to assess

statistically signiﬁcant differences within the same

group during different activities and coupling condi-

tions. Pearson correlation coefﬁcient was calculated

to evaluate the relationship between HR and SR, me-

dian ratio and median coupling parameter and age.

The results were considered statistically signiﬁcant at

P ≤ 0.05.

3 RESULTS

3.1 Participant Characteristics

Demographic and physiologic data during strong cou-

pling are summarized in table 1. The median HR dur-

ing strong coupling ranges from 87 to 115

beats

min

, with

no statistically signiﬁcant difference between the el-

derly and the young subjects. For each subject, there

are between 683375 and 1819246 HR and SR entries

acquired from Fitbit. No data from either HR or SR

are missing. In the PM reporter app, instead, the par-

ticipants registered less data, by adding from 2 to 113

training session. Moreover, only one subject did not

record any training session.

3.2 Cardiac-Locomotor Coupling

Evidence of strong coupling was found in each sub-

ject. The distribution of the ratio is not normal, but

centered around a subject speciﬁc median that ranges

between 0.87 and 1. The median of the ratio over

all subjects is 0.94. The centered distribution indi-

cates that subjects prefer to adjust their SR and HR

rather than those two quantities being independently

controlled. Figure 2 shows the distribution of the ra-

tio, plotted with different bin sizes, considering all the

subjects during the entire observational period. By re-

ducing the bin size, the peak at 1 becomes more ev-

ident. This result highlights how spontaneous strong

coupling is prevalent during normal daily activities.

BIOSIGNALS 2023 - 16th International Conference on Bio-inspired Systems and Signal Processing

172

Table 1: Demographic characteristics of the participants, heart rate and step rate during strong coupling, coupling parameter

and ratio. STD, standard deviation.

Subject Age Height Gender Heart rate Step rate Coupling parameter Ratio

Median STD Median STD Median STD Median STD

1 48 195 male 99.4 21.1 99 21.1 0.14 0.13 0.96 0.20

2 60 180 male 106 23.6 106 23.6 0.12 0.15 1.00 0.22

3 25 184 male 107 11.1 107 11.0 0.12 0.12 0.92 0.17

4 26 163 female 113 26.1 113 26.2 0.11 0.13 0.93 0.17

5 35 176 male 102 13.1 102 13.1 0.16 0.12 0.87 0.18

6 42 179 male 87.7 33.8 87 33.8 0.16 0.14 0.99 0.24

7 26 177 male 112 33.8 112 33.8 0.13 0.13 0.99 0.21

8 27 186 male 110 28.3 110 28.3 0.13 0.13 0.96 0.20

9 26 180 male 109 14.6 109 14.6 0.13 0.13 0.98 0.21

10 38 179 female 111 13.4 111 13.4 0.11 0.13 0.92 0.17

11 25 171 female 115 11.6 115 11.6 0.12 0.14 0.90 0.17

12 27 178 male 110 14.0 110 14.0 0.12 0.12 0.94 0.19

13 31 183 male 109 23.4 109 23.4 0.18 0.13 0.86 0.21

14 45 181 male 113 25.3 113 25.3 0.14 0.12 1.04 0.20

15 54 180 male 108 23.5 108 23.6 0.18 0.14 0.95 0.24

16 23 182 male 104 7.87 104 7.91 0.13 0.13 0.91 0.17

Figure 2: Ratio distribution.

The violin plots of HR and SR show that each

participant has one, two or three different step rates

at which strong coupling occur (Figure 3). This evi-

dence was used to group the subjects, as presented in

2.3.1. We observe that group 1 includes 7 subjects,

28.4 ± 5.7 years; group 2 includes 8 subjects, 39.6 ±

13.7 years, and group 3 only includes one subject, 42

years old. Overall, it can be noticed that the total per-

centage of strong coupling occurrences is higher and

the coupling parameter is smaller for the group 1 com-

pared with group 2 and 3. However, no statistically

signiﬁcance was shown in any of the above mentioned

parameters. The grey line in Figure 3 represents the

threshold for moderate intensity activity. We can no-

tice that for subjects 10 and 4, the strong coupling is

prevalent when walking at a SR of 100

steps

minute

or more.

We found higher SR in instances of strong cou-

pling vs. uncoupling (109 vs. 85, p=4.4 ∗ 10

−7

), and

in instances of weak coupling vs. uncoupling (107

vs. 85, p=0.0028). Statistically signiﬁcant differ-

ence was found also in the HR between the condi-

tion uncoupling and strong coupling (109 vs. 101, p=

6.02 ∗ 10

−4

No correlation was found between the median of

the ratio and the age or the height of the subjects. The

same was found for the median of the coupling pa-

rameter and age and height.

3.3 Timing, Duration and Intensity

The following results present the differences found in

the uncoupling, weak coupling and strong coupling

occurrences among the subjects and the activities.

We found that each subject has a percentage of

strong coupling occurrences ranging from 3% and

7% and a percentage of weak coupling occurrences

ranging from 26% and 40%, calculated over the ﬁve

months of observational period (Figure 4).

We noticed differences in the percentage of cou-

pling occurrences between the young group (aged be-

tween 23 and 40 years) and the elderly group (aged

between 40 and 60 years). Occurrences of weak cou-

pling and uncoupling are signiﬁcantly higher in the

young group (33.61 vs. 36.19, p=0.0036 and 62,02 vs.

59.07, p=0.0068, respectively). No differences were

found between the elderly group and the younger

group as regard to the strong coupling (3.98 vs. 4.74,

p=0.08).

The percentage of strong coupling occurrences

(Figure 5) was signiﬁcantly higher when perform-

Spontaneous Cardiac-Locomotor Coupling in Healthy Individuals During Daily Activities

173

Figure 3: Violin plot of heart rate and step rate for three of the subjects.

Figure 4: Coupling occurrences in percentage, mean for all

subjects.

ing moderate activities compared to light activities

(7.17% vs. 2.10%, p=5.42 ∗ 10

−6

) and when per-

forming vigorous activities compared to light activ-

ities (4.745% vs. 2.10%, p=6.01 ∗10

−4

The mean duration of activities (Figure 6) is sig-

niﬁcantly higher in presence of strong coupling events

vs. uncoupling (111.31 vs. 69.89 min, p=4.4 ∗ 10

−8

)

in strong coupling vs. weak coupling events (111.31

in vs. 80.26 min, p=0.013) and in weak coupling

events vs uncoupling (80.26 vs. 69.89 min, p=0.013).

We found no correlation between the time of the

day at which the activity is performed and the cou-

pling strength (12.89 vs. 13.10 vs. 13.00 , p=0.2 for

p=6*10

-4

p= 5* 10

-6

Figure 5: Percentage of strong coupling occurrences during

different activity intensity for each subject.

uncoupling, weak coupling and strong coupling, re-

spectively). In particular, most of the activity is dis-

tributed between 10h and 19h.

3.4 Type of Training

As regard to type of training, we found that coupling

(both strong and weak) occurred to a larger extent

during running exercises than during strength training

(6.05% vs. 4.00%, p=0.004 and 39.30% vs. 27.50%,

p=0.007, respectively).

BIOSIGNALS 2023 - 16th International Conference on Bio-inspired Systems and Signal Processing

174

p=0.013

p= 4* 10

-8

Figure 6: Mean activity duration for each subject.

4 DISCUSSION

Using lifelogging data, this study shows that every

subject presents occurrences of CLC. Furthermore,

the ratio 1:1 between SR and HR seems to occur

more frequently than other ratios. Previous studies

in laboratory settings claim that some of the subjects

never coupled (Kirby et al., 1989; Novak et al., 2007;

De Bartolo et al., 2021; Hausdorff et al., 1992). In

particular, one laboratory study found that CLC oc-

curred only in 18/25 subjects for step rates between

106 and 150

steps

minute

(Kirby et al., 1989). During daily

activities, instead, we found a median SR between 87

and 115

steps

minute

. The comparison of these two settings

indicates that the SR of the subjects in the laboratory

could be inﬂuenced by the use of the treadmill, which

might affect the choice of a comfortable walking step

rate.

When exercising, the cardiovascular system

adapts to meet the metabolic demand of the sys-

temic system including the skeletal muscles (Mur-

phy et al., 2011). Physiology tells us that the inter-

action between cardiovascular and locomotor system

originates in the interplay of the parasympathetic and

sympathetic system. During muscle activity, the sym-

pathetic nervous system is activated, which, in turn,

increases the arterial blood pressure, the HR and the

vascular resistance (Murphy et al., 2011). In this way,

we expect the HR to rise with increasing in metabolic

activity level, or increasing step frequency. The pos-

itive correlation between SR and HR was observed

for all subjects in the present study. Additionally, we

found that a 1:1 correspondence is most likely, which

raises new research questions regarding the interac-

tion of the cardiovascular and the locomotor system.

While previous studies focused only on the strong

coupling, we also investigated when and to what ex-

tent the two rhythms were within 10% to each other

(weak coupling). Even though the strong coupling is

present only in the 5% of the observational period,

during the 30% of the total time the HR and SR are

within the 10% of each other. This result highlights

that CLC can be reached without changing or forcing

the physiology of the body.

In a laboratory study, they compared CLC dur-

ing running and cycling and they demonstrated that

CLC exists for longer periods during running com-

pared to cycling (113.6s vs 58s, p < 0.05) (No-

mura et al., 2003). Our analysis shows that cou-

pling occurred to a larger extent during running com-

pared to strength trainings. Running seems to en-

hance the CLC, compared to other training activities,

which align with what we found during daily activi-

ties. Moreover, in the laboratory settings some sub-

jects coupled only while running and not while walk-

ing (Kirby et al., 1989). The differences between the

occurrence of CLC during walking compared with

running, could be due to the duration of the labora-

tory experiment. In fact, with higher exercise inten-

sity, the HR approaches the SR more rapidly than

during walking (Kirby et al., 1989). In our study,

we found that the duration of the activity is higher

in presence of strong coupling (111 min± 34 min),

whereas in the laboratory study only 2-5 min of walk-

ing were performed (Kirby et al., 1989). However,

we also found that there are some subjects who cou-

ple more while performing vigorous activities (SR >

130

steps

minute

) and other who couple more while per-

forming moderate activities (SR > 100 and < 130

steps

minute

), compared to light activities. The body seems

to approach CLC rapidly when the exercise intensity

is higher and slowly when walking at lower speed.

The extent and speed of sympathetic activation re-

sults in different HR effects depending on the subject

level of training. In a trained subject, we expect the

resting HR to be lower than in an untrained subject.

The subjects in our study have a resting HR below

beats

minute

, indicating a good level of training. During

running, the trained subject is engaged in rhythmic

exercises and we would more likely observe a sta-

ble 1:1 ratio between HR and SR. The observation

we made regarding the number of SR at which strong

coupling occur can be dependent on the training habit

of the participants, rather than on the CLC effect. The

study reports the time it took each subject to run 5km,

however, we did not ﬁnd the results very credible and

therefore did not use them in this analysis.

The small size of our study population, charac-

terized by only healthy subjects, limits any discus-

sion concerning intra-subject characteristics, like age

or height that may enhance or inhibit the CLC. In a

Spontaneous Cardiac-Locomotor Coupling in Healthy Individuals During Daily Activities

175

previous study (Novak et al., 2007), they show that

HR and SR were coupled only for the elderly group

(70.3 ± 5.1 years) and not for the young participants

(29.0 ± 5.0 years). In our study cohort, we found

no differences in the extent of strong CLC between

the participants aged <40 years vs. > 40 years old.

However, the young cohort expressed higher extent of

weak coupling. The reason could be due to the small

number of elderly subjects, which consisted of 5 sub-

jects (49.8 ± 7.2 years) in our study compared to 9

subject (70.3 ± 5.1 years) in the study (Novak et al.,

2007). Further research including more subjects with

an higher average age may be useful to investigate

age-related physiological effects.

Smartwatches are a powerful tool to obtain in-

sights and information about non-monitored daily ac-

tivities, that would otherwise be difﬁcult to obtain.

However, the advantage of unsupervised data collec-

tion could translate in non-monitored artifacts in the

data. One example could be detecting steps when the

person is using the hands for other tasks, but is not

exercising. Another could be that rhythmic move-

ment of the wrist during exercise, induces wrongly

detected heart beats. Furthermore, we expect motion

related measurement artefacts to increase with activ-

ity level. However, most coupling was found at mod-

erate intensity. The HR detection from a wrist-worn

device relies on the photoplethsmography (PPG) sig-

nal, a technique that has various limitations compared

to the chest-worn sensors, which rely directly on the

electrocardiogram (Boudreaux et al., 2018). Tight

compression of the device on the skin, changes in skin

temperature and perfusion and contraction of skeletal

muscle in the forearm and in the hand, are some of

the cause of artifacts in the PPG signal that can lead

to underestimation of HR (Boudreaux et al., 2018).

In the future, a chest-worn device with higher accu-

racy, less prone to artifacts and more robust against

HR measurements, should be utilized (Feehan et al.,

2018; Chevance et al., 2022; Sj

oberg et al., 2021).

The dataset included reports of step rates and heart

rates. However, the reports did not allow a study of

the synchronization of both time series signals with

respect to each other. In particular, it would be in-

teresting to understand if the subjects synchronized

each step with the diastolic or systolic part of the car-

diac cycle. Current wearable devices do not regis-

ter the acceleration and cardiac signals on the same

clock, which makes it harder to study such high time-

resolution phenomena. In particular, no dataset ex-

ists that records these signals during daily activities.

The identiﬁcation of CLC events during every day

activities presented in this work raises further ques-

tions to whether there is a physiologic mechanism that

controls CLC. Unfortunately, the dataset lacked suf-

ﬁcient reports of perceived exertion during coupling

and uncoupling and thus did not allow for investiga-

tion of physiological beneﬁts of CLC in terms of ex-

ercise performance. In the future it should be investi-

gated whether CLC could offer a more efﬁcient way

of training.

5 CONCLUSIONS

In conclusion, we found evidences of spontaneous

CLC during daily activities in every subject. In par-

ticular, CLC occurs more likely when the subject en-

gages in long activities at moderate intensity. More-

over, the ratio 1:1 between SR and HR seems to pre-

vail over any other ratio. By improving the under-

standing occurrence of CLC in daily life, this work

supports further research on customised training and

rehabilitation programs. Future work will address the

synchronization between cardiac contraction and the

gait cycle using temporal and spectral signal analysis

techniques.

REFERENCES

Niizeki, Kawahara, and Miyamoto (1993). Cardiac, res-

piratory, and locomotor coordination during walking

in humans. Folia primatologica., 66(1-4):226–239.

Place: [Basel, Switzerland] : Publisher: Karger.

Boudreaux, B. D., Hebert, E. P., Hollander, D. B., Williams,

B. M., Cormier, C. L., Naquin, M. R., Gillan, W. W.,

Gusew, E. E., and Kraemer, R. R. (2018). Validity of

Wearable Activity Monitors during Cycling and Re-

sistance Exercise. Medicine and Science in Sports and

Exercise, 50(3):624–633. ISBN: 0000000000.

Chevance, G., Golaszewski, N. M., Tipton, E., Hek-

ler, E. B., Buman, M., Welk, G. J., Patrick, K.,

and Godino, J. G. (2022). Accuracy and Preci-

sion of Energy Expenditure, Heart Rate, and Steps

Measured by Combined-Sensing Fitbits Against Ref-

erence Measures: Systematic Review and Meta-

analysis. JMIR mHealth and uHealth, 10(4):e35626.

Company: JMIR mHealth and uHealth Distribu-

tor: JMIR mHealth and uHealth Institution: JMIR

mHealth and uHealth Label: JMIR mHealth and

uHealth Publisher: JMIR Publications Inc., Toronto,

Canada.

Constantini, K., Stickford, A. S. L., Bleich, J. L.,

Mannheimer, P. D., Levine, B. D., and Chapman,

R. F. (2018). Synchronizing Gait with Cardiac Cy-

cle Phase Alters Heart Rate Response during Run-

ning. Medicine & Science in Sports & Exercise,

50(5):1046–1053.

De Bartolo, D., De Giorgi, C., Compagnucci, L., Betti, V.,

Antonucci, G., Morone, G., Paolucci, S., and Iosa, M.

BIOSIGNALS 2023 - 16th International Conference on Bio-inspired Systems and Signal Processing

176

(2021). Effects of cognitive workload on heart and

locomotor rhythms coupling. Neuroscience Letters,

762:136140.

Feehan, L. M., Geldman, J., Sayre, E. C., Park, C., Ez-

zat, A. M., Yoo, J. Y., Hamilton, C. B., and Li,

L. C. (2018). Accuracy of Fitbit Devices: System-

atic Review and Narrative Syntheses of Quantitative

Data. JMIR mHealth and uHealth, 6(8):e10527. Com-

pany: JMIR mHealth and uHealth Distributor: JMIR

mHealth and uHealth Institution: JMIR mHealth and

uHealth Label: JMIR mHealth and uHealth Publisher:

JMIR Publications Inc., Toronto, Canada.

Hausdorff, J. M., Forman, D. E., Pilgrim, D. M., Rigney,

D. R., and Wei, J. Y. (1992). A new technique

for simultaneous monitoring of electrocardiogram and

walking cadence. The American Journal of Cardiol-

ogy, 70(11):1064–1071.

Jett

e, M., Sidney, K., and Bl

umchen, G. (1990). Metabolic

equivalents (METS) in exercise testing, exercise pre-

scription, and evaluation of functional capacity. Clin-

ical Cardiology, 13(8):555–565.

Karami, Z., Hines, A., and Jahromi, H. Z. (2021). Leverag-

ing IoT Lifelog Data to Analyse Performance of Phys-

ical Activities. In 2021 32nd Irish Signals and Systems

Conference (ISSC), pages 1–6. ISSN: 2688-1454.

Kirby, R. L., Nugent, S. T., Marlow, R. W., Macli$od, D. A.,

Alan, Marble, E., Lee, R., Nugent, S. T., Mar, R. W.,

and Macleod, D. A. (1989). Coupling of cardiac and

locomotor rhythms. J. Appl. Physiol, pages 323–329.

Murphy, M. N., Mizuno, M., Mitchell, J. H., and

Smith, S. A. (2011). Cardiovascular regulation

by skeletal muscle reﬂexes in health and disease.

American Journal of Physiology-Heart and Circu-

latory Physiology, 301(4):H1191–H1204. eprint:

https://doi.org/10.1152/ajpheart.00208.2011.

Niizeki, K. and Miyamoto, Y. (1999). Phase-dependent

heartbeat modulation by muscle contractions dur-

ing dynamic handgrip in humans. American Jour-

nal of Physiology-Heart and Circulatory Physiology,

276(4):H1331–H1338. Publisher: American Physio-

logical Society.

Niizeki, K. and Saitoh, T. (2014). Cardiolocomotor phase

synchronization during rhythmic exercise. The Jour-

nal of Physical Fitness and Sports Medicine, 3(1):11–

20.

Nomura, K., Takei, Y., and Yanagida, Y. (2003). Compar-

ison of cardio-locomotor synchronization during run-

ning and cycling. European Journal of Applied Phys-

iology, 89(3):221–229.

Novak, V., Hu, K., Vyas, M., and Lipsitz, L. A. (2007).

Cardiolocomotor Coupling in Young and Elderly Peo-

ple. The Journals of Gerontology Series A: Biological

Sciences and Medical Sciences, 62(1):86–92.

oberg, V., Westergren, J., Monnier, A., Martire, R. L.,

Hagstr

omer, M.,

Ang, B. O., and Vixner, L. (2021).

Wrist-Worn Activity Trackers in Laboratory and Free-

Living Settings for Patients With Chronic Pain: Cri-

terion Validity Study. JMIR mHealth and uHealth,

9(1):e24806. Company: JMIR mHealth and uHealth

Distributor: JMIR mHealth and uHealth Institution:

JMIR mHealth and uHealth Label: JMIR mHealth and

uHealth Publisher: JMIR Publications Inc., Toronto,

Canada.

Takeuchi, S., Nishida, Y., and Mizushima, T. (2014). Ef-

fects of Synchronization between Cardiac and Loco-

motor Rhythms on Oxy- gen Pulse during Walking.

page 7.

Thambawita, V., Hicks, S. A., Borgli, H., Stensland, H. K.,

Jha, D., Svensen, M. K., Pettersen, S.-A., Johansen,

D., Johansen, H. D., Pettersen, S. D., Nordvang, S.,

Pedersen, S., Gjerdrum, A., Grønli, T.-M., Fredriksen,

P. M., Eg, R., Hansen, K., Fagernes, S., Claudi, C.,

Biørn-Hansen, A., Nguyen, D. T. D., Kupka, T., Ham-

mer, H. L., Jain, R., Riegler, M. A., and Halvorsen, P.

(2020). PMData: a sports logging dataset. In Proceed-

ings of the 11th ACM Multimedia Systems Conference,

pages 231–236, Istanbul Turkey. ACM.

Tudor-Locke, C., Aguiar, E. J., Han, H., Ducharme, S. W.,

Schuna, J. M., Barreira, T. V., Moore, C. C., Busa,

M. A., Lim, J., Sirard, J. R., Chipkin, S. R., and Stau-

denmayer, J. (2019). Walking cadence (steps/min)

and intensity in 21–40 year olds: CADENCE-adults.

The International Journal of Behavioral Nutrition and

Physical Activity, 16:8.

Tudor-Locke, C., Camhi, S. M., Leonardi, C., Johnson,

W. D., Katzmarzyk, P. T., Earnest, C. P., and Church,

T. S. (2011). Patterns of adult stepping cadence

in the 2005-2006 NHANES. Preventive Medicine,

53(3):178–181.

Tudor-Locke, C., Ducharme, S. W., Aguiar, E. J., Schuna,

J. M., Barreira, T. V., Moore, C. C., Chase,

C. J., Gould, Z. R., Amalbert-Birriel, M. A., Mora-

Gonzalez, J., Chipkin, S. R., and Staudenmayer, J.

(2020). Walking cadence (steps/min) and intensity in

41 to 60-year-old adults: the CADENCE-adults study.

The International Journal of Behavioral Nutrition and

Physical Activity, 17:137.

Tudor-Locke, C., Mora-Gonzalez, J., Ducharme, S. W.,

Aguiar, E. J., Schuna, J. M., Barreira, T. V., Moore,

C. C., Chase, C. J., Gould, Z. R., Amalbert-Birriel,

M. A., Chipkin, S. R., and Staudenmayer, J. (2021).

Walking cadence (steps/min) and intensity in 61-85-

year-old adults: the CADENCE-Adults study. The In-

ternational Journal of Behavioral Nutrition and Phys-

ical Activity, 18(1):129.

Spontaneous Cardiac-Locomotor Coupling in Healthy Individuals During Daily Activities

177