Chaotic Changes in Fingertip Pulse Waves during Autobiographical

Memory Retrieval

Eri Shibayama and Taira Suzuki

Graduate School of Psychology, J.F. Oberin University, Machida, Tokyo 194-0294, Japan

Keywords: Chaos, Largest Lyapunov Exponent (LLE), Autobiographical Memory.

Abstract: The effects of autobiographical memory retrieval on a psychophysiological index were examined. The

experimental (retrieval) group conducted an autobiographical memory retrieval task, whereas the control

group repeatedly vocalized Japanese syllabary. The largest fingertip pulse-wave Lyapunov Exponent (LLE)

in chaos analysis, which is a nonlinear analysis, was measured as an objective psychophysiological index.

Moreover, participants responded to a psychological questionnaire and completed an original checklist before

and after the experiment. The results indicated that fingertip pulse-wave LLE increased significantly only in

the retrieval group. This result supported previous findings that psychotherapies such as reminiscence therapy

have positive effects on cognitive functions at the psychophysiological level. Moreover, only the retrieval

group showed significant improvements in psychological indices. Therefore, recalling autobiographical

memories and verbally sharing it with others might contribute to maintaining mental health. To date,

autobiographical memory retrieval and reminiscence therapy have not been sufficiently examined by using

physiological indices. This study’s results using a physiological index are expected to contribute to research

on reminiscence therapy.

1 INTRODUCTION

Japan's population is rapidly aging, and various

countermeasures have been taken, including medical

care, welfare, and economics, among others. The

proportion of people aged 65 or older is called the

"aging rate." If a society's aging rate exceeds 21%, it

is called a super-aging society. The National Institute

of Population and Social Security Research estimates

that the aging rate is increasing, whereas the total

population is decreasing in Japan. Therefore, Japan

has become the first super-aging society globally, and

it is estimated that the aging rate will reach

approximately 30% in 2025 (Ministry of Internal

Affairs and Communications, 2013). The aging

society is an urgent issue, and problems related to

elderly adults, including medical and welfare

problems, need to be managed. Dementia, which is a

general term that covers various disorders leading to

disturbances in life, caused by a decline in intellectual

functions, is a severe health problem of older adults.

The core symptoms of dementia include memory

loss, abstract thinking and judgment disorders,

apraxia, agnosia, aphasia, and executive function

disorders.

Reminiscence therapy has been used to prevent

dementia, improve cognitive functions, and prevent

deterioration of symptoms. It is an interpersonal

assistance method and became become popular after

Butler’s publication in 1963. The basic concept of

reminiscence is that a person’s cognitive functions

improve by talking about nostalgic memories with

others. Tsuda (2015) introduced group reminiscence

therapy with older adults living in a group home for

dementia people and assessed residents and nursing

care staff's changes. The results indicated an increase

in many items in the observation scale score,

including the N-ADL. On the other hand, Tadaka et

al. (2005) reviewed previous reminiscence studies

conducted with elderly participants with dementia

and suggested that preceding studies were not always

based on a definite criterion and that the essential

effectiveness and significance of reminiscence have

not been established to date, although findings

support the efficacy of reminiscence to a certain

extent. Sufficient empirical research on the effects of

reminiscence using physiological indices has not

been conducted to date.

Shibayama, E. and Suzuki, T.

Chaotic Changes in Fingertip Pulse Waves during Autobiographical Memory Retrieval.

DOI: 10.5220/0010228500350041

In Proceedings of the 14th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2021) - Volume 4: BIOSIGNALS, pages 35-41

ISBN: 978-989-758-490-9

1.1 Chaos

At the end of the twentieth century, terms such as

“Chaos”, “Fractals”, and “Complex systems” got

much attention in the field of science. Chaos, as

academic jargon, in general terms does not mean,

“Disordered”. Chaos might be the order with a

dynamic fluctuation. According to Aihara(1993),

chaos is “a phenomenon with very complicated,

irregular, and unstable behaviors because of the

nonlinearity of the system, although the system is

following deterministic laws, and it is impossible to

predict the future state.” Chaos might be expressed as

a “fluctuation” to facilitate understanding. Many

natural and biological phenomena are complex

systems that fluctuate with chaos. For example,

fingertip pulse waves, which are biosignals, including

information of the central nervous system, and the

autonomic nervous system, among others, are

considered chaotic phenomena. Chaos analysis is a

nonlinear time series analysis that quantifies the

strength of chaos, and objectively recognizes the

psychosomatic state by extracting information that

cannot be obtained through linear analysis (Imanishi

& Oyama,2008). The fluctuation is expressed by

quantitative values such as Largest Lyapunov

Exponent (LLE), which is a quantification of

sensitivity to the initial conditions, which is one of the

characteristics of chaos. Many researchers have

indicated a correlation between a chaotic fluctuation

expressed by LLE of biological information and

mental and physical health. The external adaptability

declines, and physical and mental health cannot be

maintained when a low level of LLE continues, i.e.,

when a non-fluctuant condition continues over time.

It has been indicated that the attractor of patients with

depression or dementia has low fluctuations, and the

LLE is reduced. When depression advances, LLE

becomes further decreases (Oyama, 2012). Imanishi,

Shiomi, & Oyama (2009) and Oyama (2012) reported

that LLE of fingertip pulse waves declined when an

excessive mental and physical load was applied to

participants, suggesting that LLE of fingertip pulse

waves is correlated with mental and physical health.

The above findings suggest that LLE, which is

biological data, might be deeply correlated with

mental and physical health. The fingertip pulse wave

LLE is a useful and objective index that correlates

with mental and physical health. Moreover, fingertip

pulse waves can be measured easily using a fingertip

cuff and need only 1-2 minutes for a completed

measurement, which is a non-invasive method that

places a low physical and mental burden on

participants.

1.2 Purpose

As described above, it is necessary to use objective

indices in reminiscence studies. Efficiently assessing

reminiscence therapy's effects using objective indices

through chaos analysis of fingertip pulse waves

would contribute to further development of

reminiscence therapy and provide a useful measure

for managing different older adults' problems in a

super-aging society. However, previous studies have

not correlated reminiscence with LLE of fingertip

pulse waves. In reminiscence therapy,

autobiographical memories are shared between

clients or between the client and therapist, which is

expected to affect clients’ emotions and self-

cognition. The present study examined the effects of

recalling autobiographical memories and verbally

sharing them with others on LLE of fingertip pulse

waves. Moreover, reminiscence therapy is expected

to improve cognitive functions and relieve depression

(Arean et al., 1993). Therefore, psychological indices

were assessed before and after the experiment

examining the psychological changes resulting from

recalling autobiographical memory. In this study, we

conducted preliminary experiments on university

students, not elderly people.

2 METHODS

The experiment was conducted from October to

December of 2018. Participants (N=18) were

recruited using snowball sampling at A university

located in Tokyo. Two participants with missing data

were excluded, and the data of 16 participants were

analyzed. The mean age of female participants

(N=10) was 21.10 (SD=0.99) years, and that of male

participants (N=6) was 21.83 (SD=1.72) years. They

participated in the experiment twice; (1) conducting

an autobiographical memory recall task (retrieval

group), and (2) conducting a vocalizing Japanese

syllabary task (control group). The tasks were

performed in random order to counterbalance order

effects. The second experiment was conducted after

an interval of one or more days.

2.1 Equipment, Materials,

Psychological Scales

Fingertip pulse waves were measured as a biological

index by connecting a cuff sensor to a Lyspect 3.5

computer. The data were analyzed using a laptop

computer. The time-series data were sampled at a

frequency of 200 Hz for 180 sec. The time-delay was

set at 10.0msec, and the embedding dimension (d)

was set at 4, after Sano & Sawada (1985).

BIOSIGNALS 2021 - 14th International Conference on Bio-inspired Systems and Signal Processing

The participants also responded to the POMS 2

Japanese version for Adults (Yokoyama, 2015) and

completed an original check sheet developed for this

study after the intervention. POMS2 is composed of

seven subscales; Anger-Hostility, Confusion-

Bewilderment, Depression-Dejection, Fatigue-

Inertia, Tension-Anxiety, Vigour-Activity, and

Friendliness. Responses to POMS 2 are made using a

5-point scale ranging from 1 (Never) to 5 (Very

often). The participants were required to respond to

how they currently felt. The original check sheet

developed for this study included items inquiring

participants’ childhood memories and whether they

could talk appropriately during the experiment. The

control group participants were given a Japanese

syllabary that was printed on an A4 paper.

2.2 Procedures

First, participants were instructed, “Please choose the

most appropriate option for describing your current

feelings without thinking too deeply.” They

responded using the POMS 2 Japanese version for

Adults (Yokoyama, 2015). Then, the participants

were instructed, "We will measure your fingertip

blood flow for three minutes. Please put the cuff

around your right index finger and do not move

during the measurement. Please take a comfortable

position, breathing easily without closing your eyes,

and be natural.” After the instructions, the

participant's fingertip pulse waves at rest were

measured for three minutes (Retrieval, pre; Control,

pre). Next, the control group repeatedly read the

Japanese syllabary aloud while their finger pulse

waves were measured simultaneously (Control,

midst). The control group was instructed, “Please

slowly read aloud the Japanese syllabary written on

the paper, without considering the pronunciation or

voice volume. The measurement was conducted for

three minutes. Please restart from the beginning after

you reach the end of the syllabary.” The retrieval

group was instructed, “Please talk about the happiest

memory of your childhood. Please notify me when

you have finished describing it.” The fingertip pulse

waves were simultaneously measured (Retrieval,

midst). Next, LLE was measured again for three

minutes in both groups (Retrieval, post; Control,

post). After the measurement, the participants again

responded to POMS 2. Only the retrieval group was

requested to complete the original check sheet after

POMS 2. They were instructed, "Please check the

items that apply to you about the talk you just made

and respond to the following questions.”

2.3 Ethical Considerations

Participants were instructed that they could quit the

experiment even in the middle for any reasons and

that they would incur no disadvantages for not

responding, quitting the experiment, or based on the

content of their responses. Moreover, they were

explained that the experimental results would be

statistically processed and used only for academic

purposes and that the data would be stored in a USB

memory stick that has been encrypted. They were also

told that no individuals would be identified from the

data. The participated took part in the experiment

after giving their consent to the explanation.

2.4 Analytical Method

The original check-sheet was excluded from the

analysis. Firstly, a one-way analysis of variance

(ANOVA) was conducted to examine whether there

was a significant difference (p<.05) between the two

groups (retrieval and control) before the intervention

(pre), which indicated no significant differences.

Next, the main effects of the group, and repetition and

interactions between them were analysed.

Descriptions of the main effects are omitted, and only

significant or relatively effective interactions are

described. Finally, a simple main effects test was

conducted.

3 RESULTS

3.1 2x3 ANOVA on LLE Changes in

the Two Groups

Table 1 shows the mean LLE, standard error, standard

deviation, and 95% confidence interval for the two

groups. A 2x3 ANOVA was conducted on LLEs of

Control (pre, midst, post) and Retrieval (pre, midst,

post) groups to examine for significant differences in

the mean fingertip pulse wave LLE in

autobiographical memory recall and control groups.

The result is shown in Figure 1, indicating a

significant difference in the mean LLE between

Control midst and Retrieval midst and between

Control post and Retrieval post groups. Moreover, the

mean LLE increased in the retrieval group from pre

to midst, and from pre to post. The mean LLE

increased significantly only when recalling

autobiographical memories (Table 2).

Chaotic Changes in Fingertip Pulse Waves during Autobiographical Memory Retrieval

Figure 1: 2x3 ANOVA on LLE changes in the two groups.

3.2 ANOVA on Pre- and

Post-experiment Changes in

Psychological Indices

Table 3 shows POMS 2 fundamental statistics for

Anger-Hostility, Confusion-Bewilderment,

Depression-Dejection, Fatigue-Inertia, Tension-

Anxiety, Vigour-Activity, and Friendliness

subscales. A 2x2 ANOVA was conducted with group

and repetition as independent and POMS 2 subscale

scores as dependent variables (Table4). The results

indicated significant interactions or effective effect

sizes for all the dependent variables. Therefore, a

simple main effect test was conducted, which

indicated a significant main effect of repetition in the

retrieval group for Confusion-Bewilderment,

Depression-Dejection, Fatigue-Inertia, and Tension-

Anxiety with “post” score significantly lower than the

“pre” score in the retrieval group. These results

indicated that depression, dejection, fatigue, inertia,

tension, and anxiety decreased significantly only in

the retrieval group.

Table 1: Basic statistics.

Table 2: ANOVA of the mean LLE for both groups.

=16

Mean SD SE

95%CI Lower 95%CI Upper

Control pre 4.39 1.25 0.31 3.73 5.06

Control midst 4.35 1.02 0.25 3.81 4.89

Control post 4.10 0.90 0.23 3.62 4.58

Retrieval pre 4.31 1.20 0.30 3.67 4.95

Retrieval midst 6.22 1.07 0.27 5.65 6.80

Retrieval post 5.19 1.60 0.40 4.33 6.04

Fpη²

Result・Multiple comparisons(Holm)

Interaction 10.80(2,60) <.001 .21

Main effect pre .04(1,30) .85 .00

Main effect midst 25.69(1,30) <.001 .46 Control＜Retrieval

Main effect post 5.61(1,30) .03 .16 Control＜Retrieval

Simple main effect(Contlol) .80(2,30) .46 .05

Simple main effect(Retrieval) 15.96(2,30) <.001 .52 pre<midst(

(14)=5.35,

=.00,

=1.34)

pre<post(

(14)=3.32,

=.01,

=.83)

midst>post(

(14)=2.71,

=.02,

=.68)

BIOSIGNALS 2021 - 14th International Conference on Bio-inspired Systems and Signal Processing

3.2 Anova on Pre- and

Post-experiment Changes in

Psychological Indices

Table 3 shows POMS 2 fundamental statistics for

Anger-Hostility, Confusion-Bewilderment,

Depression-Dejection, Fatigue-Inertia, Tension-

Anxiety, Vigour-Activity, and Friendliness

subscales. A 2x2 ANOVA was conducted with group

and repetition as independent and POMS 2 subscale

scores as dependent variables (Table4). The results

indicated significant interactions or effective effect

sizes for all the dependent variables. Therefore, a

simple main effect test was conducted, which

indicated a significant main effect of repetition in the

retrieval group for Confusion-Bewilderment,

Depression-Dejection, Fatigue-Inertia, and Tension-

Anxiety with “post” score significantly lower than the

“pre” score in the retrieval group. These results

indicated that depression, dejection, fatigue, inertia,

tension, and anxiety decreased significantly only in

the retrieval group.

Table 3: POMS 2 fundamental statistics.

Table 4: Results of analysis of variance of psychological indicators.

=16

Mean SD SE

95%CI Lower 95%CI Upper

Anger-Host ility

7.44 4.26 1.06 5.17 9.71

Confusion-Bewilderment

9.81 4.34 1.08 7.50 12.12

Depression-Dejection

8.38 4.57 1.14 5.94 10.81

Fatigue-Inertia

10.25 4.11 1.03 8.06 12.44

Tension-Anxiety

9.81 3.87 0.97 7.75 11.87

Vigour-Activity

11.88 5.52 1.38 8.93 14.82

Friendliness

16.38 3.93 0.98 14.28 18.47

=16

Mean SD SE

95%CI Lower 95%CI Upper

Anger-Host ility

6.19 2.07 0.52 5.08 7.29

Confusion-Bewilderment

6.19 4.14 1.03 3.98 8.39

Depression-Dejection

8.25 4.89 1.22 5.64 10.86

Fatigue-Inertia

9.94 3.43 0.86 8.11 11.77

Tension-Anxiety

9.50 3.43 0.86 7.67 11.33

Vigour-Activity

11.06 5.13 1.28 8.33 13.80

Friendliness

15.63 3.76 0.94 13.62 17.63

=16

Mean SD SE

95%CI Lower 95%CI Upper

Anger-Host ility

7.88 4.30 1.08 5.58 10.17

Confusion-Bewilderment

11.06 3.40 0.85 9.25 12.87

Depression-Dejection

8.63 2.36 0.59 7.37 9.88

Fatigue-Inertia

10.44 3.24 0.81 8.71 12.17

Tension-Anxiety

9.88 3.26 0.82 8.14 11.61

Vigour-Activity

12.38 4.63 1.16 9.91 14.84

Friendliness

17.56 4.27 1.07 15.29 19.84

=16

Mean SD SE

95%CI Lower 95%CI Upper

Anger-Host ility

6.38 1.82 0.46 5.40 7.35

Confusion-Bewilderment

8.69 3.16 0.79 7.01 10.37

Depression-Dejection

7.38 2.22 0.55 6.19 8.56

Fatigue-Inertia

8.94 3.11 0.78 7.28 10.59

Tension-Anxiety

8.25 2.35 0.59 7.00 9.50

Vigour-Activity

11.94 4.58 1.15 9.50 14.38

Friendliness

16.56 4.70 1.18 14.06 19.07

Retrieval(post)

Control(post)

Control(pre)

Retrieval(pre)

Fpη²

Interaction 18.63(1,30) <.001 .31

Main effect pre .82(1,30) .37 .03

Main effect post 1.12(1,30) .30 .04

Simple main effect(Contlol) .60(1,15) .45 .04

Simple main effect(Retrieval) 21.24(1,15) <.001 .59

Interaction 4.55(1,30) .40 .11

Main effect pre .04(1,30) .85 .00

Main effect post .43(1,30) .52 .01

Simple main effect(Contlol) .32(1,15) .58 .02

Simple main effect(Retrieval) 6.82(1,15) .02 .31

Interaction 2.33(1,30) .14 .06

Main effect pre .02(1,30) .89 .00

Main effect post .75(1,30) .40 .02

Simple main effect(Contlol) .46(1,15) .51 .03

Simple main effect(Retrieval) 5.75(1,15) .03 .28

Interaction 3.36(1,30) .08 .08

Main effect pre .00(1,30) .96 .00

Main effect post 1.45(1,30) .24 .05

Simple main effect(Contlol) .38(1,15) .55 .03

Simple main effect(Retrieval) 10.26(1,15) .01 .41

Confusion

-Bewilderment

Depression

-Dejection

Fatigue

-Inertia

Tension

-Anxiety

Chaotic Changes in Fingertip Pulse Waves during Autobiographical Memory Retrieval

4 DISCUSSION

The results indicated that participants' psychological

conditions only improved when recalling

autobiographical memories. Moreover, the mean

LLE increased significantly, which suggested that

fingertip pulse wave LLE significantly increases

when recalling autobiographical memories and

verbally sharing them with others.

As described above, fingertip pulse waves,

which are biosignals, are considered chaotic

phenomena that include information about the central

nervous system and the autonomic nervous system.

Therefore, chaotic pulse wave phenomena include

physiological signals and information related to

psychological conditions, such as feelings, emotions,

and cognitive functions. Suzuki & Suzuki (2007)

measured fingertip pulse waves of older adults with

dementia before and after painting a picture, which

indicated that the dynamic structure of chaos

attractors was significant after their cognitive

functions were stimulated by painting. Moreover,

Oyama (2012) indicated that LLE decreased when

communication skills decreased or dementia

progressed, and low LLE continued depressed

people. Consistently with the previous studies, the

present study indicated that participants’ LLE

increased significantly when their psychological

condition turns positive as a result of recalling and

sharing autobiographical memories.

The results of this study suggest that

psychological indices can be improved by recalling

autobiographical memories. Takahashi & Matsuo

(2007) examined pre- and post- autobiographical

memory recall changes in 69 university women's

emotions and reported a significant decrease in post-

recall negative emotions. Moreover, Kobayashi,

Iwanaga ,& Ikuwa (2002) suggested that

autobiographical memories would be strongly

recalled when feeling very nostalgic, and participants

would develop positive feelings. The above studies

have suggests that mental health can be maintained

by interventions using autobiographical memory

retrieval, which was supported by the present study.

Based on the above findings, it is suggested that

the effects of psychotherapy, including reminiscence,

i.e., recalling autobiographical memories and

verbally interacting with others, can be measured by

using LLE. The fingertip pulse wave measurement

used in the current study is a non-invasive method

that places a low physical and mental burden on

participants because the equipment is easy to use, and

the measurement needs only a short time. Although

wearing a cuff around the fingertip might disturb

daily life movements, pulse waves can also be

measured at the earlobes. Empirical research using

physiological indices are required for examining the

effects of reminiscence. Using fingertip pulse waves

LLE as an index could facilitate the easy assessment

of reminiscence's temporary and long-term effects. In

the future, different physiological indices should be

used to assess the effects of reminiscence.

This study examined the effects of

autobiographical memory retrieval on fingertip pulse

waves LLE in healthy university students. However,

it is unknown if the results would be similar when

using reminiscence with older adults or dementia

patients. In the future, the correlation between

communication skills or cognitive functions and

fingertip pulse waves LLE should be examined with

older adults and patients with dementia to examine

the effectiveness of fingertip pulse waves LLE as an

objective index of cognitive functions. Moreover, the

control group's task in the present study was

vocalizing Japanese syllabi, which is rather

monotonous. In the future, chaotic fluctuations in

fingertip pulse waves LLE when talking about topics

that are not associated with autobiographical memory

retrieval should be compared to examine the meaning

and functions of reminiscence in detail. The

accumulation of such studies would contribute to the

development of research on psychotherapy effects,

including reminiscence associated with

autobiographical memory retrieval. Also, studies on

dementia, improvement in cognitive functions, and

prevention of symptom deterioration could benefit

from such research.

REFERENCES

Ministry of Internal Affairs &Communications(2013).

WHITE PAPER Information and Communications in

Japan 25.

Rieko Tsuda(2015)．Effect of Implementing Group Life

Review in a Group Home ， Social Welfare ，

56(2),141-151.

Etsuko Tadaka et. al. (2005) ．Literature Review on the

Significance and Effect of Reminiscence on the Elderly

with Dementia ， Journal of Japan Academy of

Geromtological Nursing，9(2),56-63.

Kazuyuki Aihara(1993)．CHAOS A completely new wave

of creation-. Tokyo, Koudansya.

Akira Imanishi Oyama(2008)．Proposal of new analysis

method in physiological psychology - Chaotic Analysis

of Biological Signals -，Humanities review,58(3),23-

42.

Mayumi Oyama (2012) ．Psychology of “fluctuation” for

strengthening the mental immune system ， Tokyo,

Shodensha．

Akira Imanishi ， Kakuichi Shiomi ， & Mayumi

Oyama(2009). Psychophysiological evaluations of the

BIOSIGNALS 2021 - 14th International Conference on Bio-inspired Systems and Signal Processing

truck- driving for 9-hr. at night: An analysis by the

Lyapunov exponent of driver’s ear-plethysmograms,

The 73rd Annual Convention of the Japanese

Psychological Association，1338．

Arean, P. A., Perri, M. G., Nezu, A. M., Schein, R. L.,

Christopher, F., & Joseph, T. X. (1993) ． Comparative

effectiveness of social-solving therapy and

reminiscence therapy as treatments for depression in

older adults. Journal of consulting and clinical

psychology, 61, 1003-1010.

Kouji Sano & Yasuji Sawada (1985)．Measurement of the

Lyapunov spectrum from a chaotic time series ，

Physical Review Letters，55,1082-1085.

Kazuhito Yokoyama(2015)． Profile of Mood States 2nd

Edition Japanese Version，Kanekosyobou.

Takako Suzuki & Taira Suzuki(2007) ． Examination of

digital plethysmogram in elderly patients with dementia

-Changes in chaotic attractors of blood flow before and

after painting with paints- (in Japanese) ，The 71rd

Annual Convention of the Japanese Psychological

Association，1135．

Mina Takahashi & Takanori Matsuno (2007)．Effects of

recollecting important and fun autobiographical

memories on mood and the sense of self-affirmation，

Annual Bulletin of Institute of Psychological Studies

Showa Women’s University, 19, 59-69.

Kobayashi Iwanaga Ikuwa(2002)．Relations of nostalgia

with music to emotional response and recall of

autobiographical memory，Memoirs of the Faculty of

Integrated Arts and Sciences 4,28,21-28.

Chaotic Changes in Fingertip Pulse Waves during Autobiographical Memory Retrieval