ArticlePDF Available

Abstract and Figures

A copy of the article can be downloaded from my website publications.lilianjansbeken.nl. The purpose of this review is to extend previous review findings by providing an updated overview of the literature on the connection of gratitude to human health, specifically focusing on experimental study findings, to better understand possible causation, complemented with findings from multi-wave longitudinal studies.
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
https://www.tandfonline.com/action/journalInformation?journalCode=rpos20
The Journal of Positive Psychology
Dedicated to furthering research and promoting good practice
ISSN: 1743-9760 (Print) 1743-9779 (Online) Journal homepage: https://www.tandfonline.com/loi/rpos20
Gratitude and health: An updated review
Lilian Jans-Beken, Nele Jacobs, Mayke Janssens, Sanne Peeters, Jennifer
Reijnders, Lilian Lechner & Johan Lataster
To cite this article: Lilian Jans-Beken, Nele Jacobs, Mayke Janssens, Sanne Peeters, Jennifer
Reijnders, Lilian Lechner & Johan Lataster (2019): Gratitude and health: An updated review, The
Journal of Positive Psychology, DOI: 10.1080/17439760.2019.1651888
To link to this article: https://doi.org/10.1080/17439760.2019.1651888
Published online: 06 Aug 2019.
Submit your article to this journal
View Crossmark data
Gratitude and health: An updated review
Lilian Jans-Beken
a
, Nele Jacobs
a,b
, Mayke Janssens
a,b
, Sanne Peeters
a,b
, Jennifer Reijnders
a
,
Lilian Lechner
a
and Johan Lataster
a,b
a
Faculty of Psychology and Educational Sciences, Open University, Heerlen, The Netherlands;
b
Department of Psychiatry and Psychology,
School for Mental Health and Neuroscience, Maastricht University Medical Centre, Maastricht, The Netherlands
ABSTRACT
The purpose of this review is to extend previous review ndings by providing an updated
overview of the literature on the connection of gratitude to human health, specically focusing
on experimental study ndings, to better understand possible causation, complemented with
ndings from multi-wave longitudinal studies.
ARTICLE HISTORY
Received 16 January 2019
Accepted 19 July 2019
KEYWORDS
Gratitude; health; physical
health; mental health
The study of gratitude, perceived as an important
source of human strength, has gained increasing atten-
tion over the past decades. Around 2010, several
reviews appeared that evaluated the contribution of
gratitude to mental and physical health (Emmons &
Mishra, 2011; Wood, Froh, & Geraghty, 2010). While
the authors of these reviews consistently concluded
that gratitude links positively to positive emotions and
subjective well-being, and negatively to emotional vul-
nerabilities and negative aect, most studies in these
reviews used cross-sectional observational designs,
leaving causality of relationships unclear. In addition,
the eects of gratitude on physical health had until
then, been left virtually unexplored. The aim of the
current study is, therefore, to extend earlier review
ndings by providing an updated overview of the lit-
erature on the connection of gratitude to human
health, specically focusing on experimental study nd-
ings, to gain better insight in possible causation, com-
plemented with ndings from multi-wave longitudinal
studies. Findings presented in this review will provide
guidance to scholars and practitioners about the scien-
tic study of gratitude and its translation into practice.
Gratitude
Scientists conceptualize gratitude as both a state and
a trait. State gratitude is an attribution-dependent or
aective-cognitive state based on the ability to be
empathic, resulting from both appraising a received ben-
et as a positive outcome as well as recognizing that this
positive outcome stems from an external source. The
grateful emotion promotes (upstream) reciprocity, and
prosocial behaviour (Bartlett & DeSteno, 2006;Clore,
Ortony, & Foss, 1987; Lazarus & Lazarus, 1996; Nowak &
Roch, 2007; Tsang, 2006; Weiner, 1985; Wood, Maltby,
Stewart, & Joseph, 2008). Trait gratitude can be viewed
as a wider life orientation towards noticing and being
grateful for the positive in the world. Attention can be
directed to the feeling of suciency, to the appreciation
of the little things in life, and to other people in our lives
(Thomas & Watkins, 2003). Individuals with a grateful per-
spective on life are more likely to show (pro)social beha-
viours (Wood et al., 2010), theorized to at least partly
underly previously established associations between gra-
titude and health-related outcomes. The results, suggest-
ing state and trait gratitude being benecial for physical
and mental health, have led to the development of grati-
tude interventions to decrease psychological symptoms
and increase physical and mental well-being.
Gratitude interventions
A variety of gratitude interventions are used to induce or
increase levels of gratitude, often with the aim to reduce ill-
being and improve well-being. Commonly used interven-
tions to increase levels of gratitude are gratitude journaling,
writing a gratitude letter, and the Three Good Things (TGT)
exercise. Gratitude journaling consists of writing on
a regular basis about things, people, and events one feels
explicitly grateful for. The frequency of writing diers
between studies, ranging from writing a single time to
daily (DeWall, Lambert, Pond, Kashdan, & Fincham, 2012;
Flinchbaugh, Moore, Chang, & May, 2012; Jackowska,
Brown, Ronaldson, & Steptoe, 2016; Kerr, ODonovan, &
Pepping, 2015). The gratitude letter is part of the gratitude
visit as devised by Seligman, Rashid, and Parks (2006).
CONTACT Lilian Jans-Beken info@lilianjansbeken.nl
THE JOURNAL OF POSITIVE PSYCHOLOGY
https://doi.org/10.1080/17439760.2019.1651888
© 2019 Informa UK Limited, trading as Taylor & Francis Group
Usually, the respondent addresses the letter to someone
they are grateful for in life, but who they never properly
thanked. After composing the letter, the content is read
aloud to the intended recipient; however, in most experi-
ments, this letter remains undelivered. The TGT exercise
(Seligman, Steen, Park, & Peterson, 2005) is similar to gra-
titude journaling, except that the instruction is to write
down three good things that happened in a specied
period, ranging from once a day to once a week (Chan,
2011;Krentzmanetal.,2015). Lastly, several experimental
setups have been used to induce a state of gratitude in
a laboratory context (Peters, Meevissen, & Hanssen, 2013;
Yu,Cai,Shen,Gao,&Zhou,2016). All these interventions
aim to increase or improve state or trait gratitude.
Health
The World Health Organization denes health since
1948 as a state of complete physical, mental and social
well-being and not merely the absence of disease or
inrmity. The denition has been increasingly criticized
for being impracticable and counterproductive in an
era where ageing with chronic illnesses has become
the norm, thereby contributing to medicalization of
society. Therefore, Huber et al. (2011) have proposed
a new general concept of health: Health as the ability
to adapt and to self-manage, in the face of social,
physical and emotional challenges, and coined it posi-
tive health. This general concept represents a broader
view on health as a dynamic ability to adjust to lifes
challenges with resilience, and to self-manage ones
own well-being. Although the denitions dier, they
both recognize that health broadly consists of
aphysical and a mental component.
Physical health
Physical health refers to the well-functioning body, the
processes therein, and the perception of physical t-
ness. External factors, the intrusion of viruses or bac-
teria, or malfunction from processes such as digestion
or respiration can harm the body. Physiological pro-
cesses such as sleep or pain can be disturbed, causing
physical constraints or deterioration of subjectively per-
ceived physical tness.
Mental health
Mental health, according to the dual-continua model of
Keyes (2002,2005), is represented along two dimensions:
the absence or presence of psychopathological symp-
toms, and the absence or presence of well-being. The
rst continuum speaks for itself; the individual indicates
the level of psychopathological symptoms experienced,
such as symptoms of depression or anxiety. The well-
being continuum encompasses emotional, psychologi-
cal, and social components of well-being. Characteristics
of emotional well-being are feelings of happiness joy,
and contentment (Diener, 2000). Psychological well-
being focuses on, for instance, the experience of auton-
omy, competence, and meaning in life (Ry,2014). Social
well-being is, among others, about relationships with
others, feeling accepted by others, and belonging
(Keyes, 2002,2005). The abovementioned physical and
mental well-being components together form a holistic
concept of human health.
This literature review will provide an updated over-
view of existing gratitude research related to compo-
nents of the dierent dimensions of human health:
physical health, psychopathology, emotional well-
being, psychological well-being, and social well-being.
The focus of this literature review will be on experimen-
tal studies with state and trait gratitude as independent
variable and their association with health-related con-
cepts, to gain insight in possible causality between gra-
titude and human health, complemented with ndings
from longitudinal studies. Such an overview can reveal
gaps in current research that can guide future scientic
research and support the choice of gratitude interven-
tions in practice.
Method
PsycINFO and PubMed databases were screened to
obtain articles from the elds of psychology and medi-
cine, using PRISMA guidelines to report on the search
ndings (Moher, Liberati, Tetzla, & Altman, 2009), see
Figure 1. The most recent reviews on gratitude date from
20102011 (Emmons & Mishra, 2011; Wood et al., 2010),
and the current review therefore focused on articles
published from the 1st of January 2010 until the 31st of
July 2018. Only articles from international, peer-reviewed
academic journals were included to ensure academic
quality. As our study aimed to move beyond correla-
tional evidence, focus was on experimental studies, com-
plemented with longitudinal studies with at least two
waves of measurement. Wood et al. (2008) showed in
their research that the Gratitude Questionnaire (GQ6:
McCullough, Emmons, & Tsang, 2002), the subscales of
the Short Gratitude, Resentment, and Appreciation Test
(SGRAT: Thomas & Watkins, 2003), and the subscales of
the Appreciation Scale (Fagley & Adler, 2012) all pertain
to the same latent gratitude construct. We therefore
followed their advice to incorporate both state and
trait gratitude, to ensure inclusion of a broad range of
studies. Additionally, as gratitude and appreciation are
used interchangeably in the scholarly literature, we were
also interested in studies examining appreciation and
2L. JANS-BEKEN ET AL.
health. To obtain articles reporting on the results of
quantitative longitudinal observational and intervention
studies with gratitude as predictor of health-related out-
comes, we used the following search terms for the title:
gratitude,grateful,thankful, and appreciation.
Search terms for the abstract were experiment,inter-
vention,prospective, and longitudinal. Combinations
of these search terms were used in sixteen (four x four)
search commands. Articles fullling search criteria were
rst screened based on the information in the abstract.
Articles with other designs than experimental, interven-
tional, prospective, or longitudinal studies with gratitude
as dependent variable, and studies unrelated to state or
trait gratitude and/or health were excluded, leaving 64
studies eligible for review (Figure 1). If eect sizes were
not reported in the original paper, they were estimated
based on available data, using the method reported by
Lakens (2013).
Review
Physical health concepts
The rst component of health consists of physical out-
comes. The articles found, report on bodily functions,
pain, and other physical complaints, perceived physical
health, and overall tness. Our search identied nine
studies reporting on the eects of experimentally
induced gratitude on (i) cardiovascular physiology, (ii)
biomarkers for stress and inammation, (iii) pain per-
ception, and (iv) sleep, and two longitudinal observa-
tional studies on the prospective eects of gratitude on
(v) (perceived) physical health. Table 1 presents an over-
view of the included articles.
Cardiovascular physiology
Randomized controlled trials (RCTs) on the relationship
between gratitude and cardiovascular physiology by
Rash, Matsuba, and Prkachin (2011), Jackowska et al.
(2016), and Redwine et al. (2016) have yielded mixed
results. Keeping a gratitude journal, in the study by
Jackowska et al. (2016), did not benecially aect heart
rate nor systolic blood pressure compared to everyday
events recall, although diastolic blood pressure revealed
to drop signicantly after gratitude journaling in compar-
ison to no-treatment conditions. Redwine et al. (2016)did
not nd dierences in heart rate variability (HRV) at rest in
a sample of heart disease patients when comparing
a gratitude intervention and treatment as usual group,
although researchers observed increased parasympa-
thetic HRV in the intervention group. Rash et al. (2011)
did, however, report a higher degree of cardiac coher-
ence suggested to reect increased physiological coor-
dination following gratitude contemplation compared
to memorable event recall.
Biomarkers for stress and inammation
In a sample of patients with heart disease, Redwine
et al. (2016) extracted a selection of inammatory bio-
markers from blood (CRP, TNF-α, IL-6, and sTNFr1),
before and after an 8-week gratitude journaling inter-
vention. Overall biomarker concentrations reduced mar-
ginally but signicantly in the gratitude intervention
compared to the treatment as usual group. Marginal
eects of gratitude on blood-based tumour necrosis
factor-α(TNF-α), but not on other inammatory biomar-
kers, were also reported in an observational prospective
study in post-acute coronary syndrome patients
Figure 1. Number of articles found in PsycINFO and PubMed with indicated search terms using lters year 20102017, academic
journal articles and English language.
THE JOURNAL OF POSITIVE PSYCHOLOGY 3
Table 1. Physical health: summary of articles.
First author
(year)
Study
design
N
(condition) Type participant
%
Male Mean Age (SD)
Gratitude
measure(s) Dependent variable(s) Intervention
Time
frame ES Summary of ndings
Baxter (2012) E 4(EG)
4(ACG)
Adults with chronic
back pain
50 55 (8.25) . . . Pain Character strength
and gratitude
intervention
57 weeks Not able to calculate The Character strength and
gratitude intervention
had no eect on pain.
Digdon (2011) E 13 (EG)
13 (ACG)
11 (ACG)
College students 22 23 (6.11) . . . Sleep quality; Pre-sleep
worry; arousal
Positive events
journaling (end
of day)
1 week d= .63 (p< .01)
d= .60 (p< .001)
Reduction of pre-sleep
arousal, improving sleep
quality and duration, no
eects on bedtime
thinking, planning or
anxiety, and sleep onset
latency.
Human(2015) L 164 (T) Older adults after
acute coronary
syndrome (ACS)
84 62 (10.60) TG (GQ6) Physical activity
(Accelerometer);
Inammatory
biomarkers; Cardiac
readmission
. . . 6 months Not able to calculate
β= .009 (p< .05)
No eects of baseline
gratitude on post-test
physical activity,
rehospitalization, and
marginal benecial eect
of gratitude on tumor
necrosis factor-α(TNF-α).
Jackowska (2016) E 40 (EG)
41 (ACG)
38 (CG)
Young adults 0 26 (0.77)
27 (0.79)
26 (0.82)
. . . Blood pressure; Heart
rate variability;
Cortisol; Sleep quality
Gratitude journal (3x
a week for 2
weeks)
4 weeks Not able to calculate Sleep quality improved in
the gratitude condition;
No changes for blood
pressure, heart rate or
cortisol in gratitude
condition.
Millstein (2016) L 156 (T) Older adults after
acute coronary
syndrome (ACS)
84 62 (10.60) TG (GQ6) Physical health .. . 2 weeks post-ACS
and 6 month
follow-up
β= .09 (n.s.)
β= .01 (n.s.)
No association was found
between gratitude and
measures of physical
health, adjusted for
baseline values, gender,
age, race, medical and
social risk factors, and
anxiety and depression.
Rash (2011) E 56 (T)
NR (EG)
NR (ACG)
Adult sample 54 23 (3.00) TG (GQ6); ST
(thinking of
grateful
things)
Cardiac coherence Gratitude
contemplation
(twice a week)
4 weeks η
2
= .14 (p< .05) Cardiac coherence during
the gratitude induction
was signicantly higher
than during the
memorable events
induction.
Redwine
(2016)
E 24 (EG)
34 (ACG)
Older patients with
stage B heart
failure
90 66 (7.58) TG (GQ6) Inammatory
biomarkers; Heart rate
variability
Gratitude journal
(daily)
8 weeks η
2
= .21 (p< .01)
η
2
= .14 (p< .05)
η
2
= .12 (n.s.)
Reduction of inammatory
biomarker index and
increased parasympathic
heart rate variability in
gratitude condition; No
change in resting heart
rate variability.
Schnitker (2018) E 33 (normal)
37 (social)
31 (prayer)
Graduate students 12 18 (0.42) TG (GQ6) Health Dierent kinds of
gratitude
journaling once
a week for 5
weeks
6 weeks Not able to calculate None of the gratitude
journaling versions
improved reported
health.
(Continued)
4L. JANS-BEKEN ET AL.
Table 1. (Continued).
First author
(year)
Study
design
N
(condition) Type participant
%
Male Mean Age (SD)
Gratitude
measure(s) Dependent variable(s) Intervention
Time
frame ES Summary of ndings
Southwell (2017) E 75 (EG)
52 (CG)
Adults with
depression and/or
anxiety
12 34 (10.80) TG (GQ6) Sleep quality Daily diary for at
least 3 x per week
for 3 weeks
9 weeks η
2
= .14 (p= .015) Sleep quality only increased
between pretest and
posttest, not posttest
and follow-up or pretest
and follow-up.
Yu
(2016)
E 15 (T) College students 20 21
(0.75)
. . . Perceived pain intensity;
Gratitude towards
partner
Sharing pain
induction with
partner who
decided or was
forced to help
... η
2
= .28 (p< .05)* Intentional help was
associated with lower
perceived pain intensity.
Yu
(2016)
E 27 (T) College students 41 22
(1.5)
TG (GQ6) Allocation of money
points: Gratitude
towards partner
Sharing pain
induction with
partner who
decided or was
forced to help in
an fMRI scanner
... η
2
= .47 (p< .001)* Intentional help was
associated with higher
money points allocation.
Several brain regions
could be appointed that
were involved in this
reciprocity elicited by
gratitude.
Note. E = experimental; T = total group, EG = experimental group; CG = control group; ACG = active control group; NR = not reported; TG = trait gratitude; SG = state gratitude; GQ6 = Gratitude Questionnaire 6; GAC =
Gratitude Adjectives Checklist, ES = eect size, * = estimation based on results in article.
THE JOURNAL OF POSITIVE PSYCHOLOGY 5
(Human et al., 2015). Jackowska et al. (2016) did not
nd evidence in their RCT for changes in salivary corti-
sol measures as a result of keeping a gratitude diary
versus both active and no-treatment control conditions.
Pain perception
In a study by Yu et al. (2016), healthy college students
were exposed to a pain induction experiment, in which
they interacted virtually with an anonymous partner
that either intentionally (gratitude condition) or unin-
tentionally bore part of their pain. Participants had to
rate their perceived pain intensity and interpersonal
closeness toward the partner, and/or express reciprocity
by transferring an amount of money. Pain was per-
ceived as less intense when receiving help was inter-
preted as intentional, relative to unintentional. A small
pilot study with cross-over multi-baseline design by
Baxter, Johnson, and Bean (2012), on the other hand,
did not show any eect of a gratitude intervention on
pain perception in people with chronic back pain.
Sleep
The gratitude intervention study by Jackowska et al.
(2016) included assessment of sleep quality and sleep
disturbance. Daily sleep quality improved to a slightly,
but signicantly greater extent following 2 weeks of
gratitude journaling compared to no-treatment control
conditions. Also, in a pre-post-follow-up trial (Southwell
& Gould, 2017), sleep quality improved moderately from
pre- to post-test, but not at follow-up, in the gratitude
journaling intervention vs. waitlist control condition.
However, no dierences in changes in sleep quality
nor sleep disturbances were found between the grati-
tude intervention and active control (everyday events
recall) group: subjective sleep ratings improved equally
in both groups (Jackowska et al., 2016). A randomized
pilot trial by Digdon and Koble (2011) has suggested
that focusing on something positive for a brief period
each evening (gratitude intervention) reduces pre-sleep
arousal, as well as improving sleep quality and duration,
but not more so than when engaging in constructive
worry or imagery distraction exercises.
Physical health outcomes
Two studies investigated prospective associations
between gratitude measured 2 weeks after acute coron-
ary syndrome (ACS), and physical health outcomes 6
months later (Human et al., 2015; Millstein et al., 2016).
Gratitude did not predict physical health-related quality of
life, physical functioning status (Millstein et al., 2016),
objectively measured physical activity or rehospitalization
(Human et al., 2015). In a study with gratitude framed as
prayer, participants did not report increased health over
the course of 6 weeks (Schnitker & Richardson, 2018).
Millstein et al. (2016), however, observed a positive eect
of gratitude on self-reported adherence to cardiac health
behaviours, the rst factor indirectly linked to physical
health in previous research (Lamers, Bolier, Westerhof,
Smit, & Bohlmeijer, 2012; Lavelock et al., 2016), and the
latter directly associated with reduced morbidity and
mortality after ACS (Chow et al., 2010).
In conclusion
The growing body of prospective and experimental
work on the eects of gratitude on bodily functions
has so far produced inconclusive results. On the one
hand, gratitude interventions appear to positively aect
a number of cardiovascular and inammatory para-
meters, as well as improving sleep quality. On the
other hand, the eects of gratitude exercises on bodily
functions do generally not distinguish from those of
other recall or distraction exercises, underlining the
need for further research to clarify to which specic
and/or generic intervention aspects these eects can
be attributed. Lastly, there is currently no convincing
evidence to support a causal link between gratitude
and (reduced) pain perception, and gratitude does not
seem to directly predict physical health outcomes when
examined prospectively.
Psychopathology concepts
Following the dual-continua model of mental well-
being of Keyes (2002,2005) the next paragraph con-
tains articles concerning indicators of cognitive and
emotional (dis)functioning. Our search yielded 25
experimental and 9 prospective observational studies
on the relationship between gratitude and (i) depres-
sion, (ii) anxiety, (iii) stress, (iv) negative aect, (v) other
psychopathological symptoms, and (vi) aggression.
Table 2 presents an overview of included studies.
Depression
Findings from RCTs in healthy samples, across a wide age
range, have suggested a variety of gratitude interventions
to moderately reduce levels of depression immediately
after the intervention (Cheng, Tsui, & Lam, 2015;
Jackowskaetal.,2016;OConnell, OShea, & Gallagher,
2017b; Ramírez, Ortega, Chamorro, & Colmenero, 2014;
Salces-Cubero, Ramírez-Fernández, & Ortega-Martínez,
2018; Watkins, Uhder, & Pichinevskiy, 2015;Wolfe&
Patterson, 2017), at approximately 1-month follow-up
(Salces-Cubero et al., 2018), and at 3-months follow-up
(Cheng et al., 2015;OConnell et al., 2017b; Ramírez et al.,
2014). Corroborating support for benecial eects of gra-
titude interventions on depression comes from
6L. JANS-BEKEN ET AL.
Table 2. Psychopathology: summary of articles.
First
author (year)
Study
design
N
(condition) Type participant % Male
Mean Age
(SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
Baxter
(2012)
E 8 (T)
4(EG)
4(ACG)
Adults with chronic
back pain
50 55 (8.25) .. . Anger; Sadness;
Anxiety;
Depression
Character strength and
gratitude
intervention
57 weeks Not able to calculate The Character strength and
gratitude intervention
decreased anger but there
was no eect on sadness,
anxiety, or depression.
Chan (2011) E 63(T) Chinese school
teachers
16 34 (6.91) TG (GQ6) Burnout Weekly log of TGT and
Naikan questions
8 weeks d= .38 (p< .05) The intervention decreased
emotional exhaustion and
depersonalization in the
high meaningful-life group
Cheng
(2015)
E 34 (EG)
34 (ACG)
34 (CG)
Health care workers 35 NR . . . Depression;
Perceived stress
Gratitude journal
(daily for 4 weeks)
4 months d=.49 (p< .05)
d=.70 (p< .01)
Perceived stress and
depression decreased after
three months at follow up
but the rate of the decline
became less obvious as the
time progressed.
Deng (2018) E 29 (EG)
37 (ACG)
30 (CG)
Male prisoners 100 35 (9.65) TG (GRAT) Aggression Daily diaries (counting
blessing) versus 5
weekly group
sessions (sharing
gratitude)
5 weeks η
2
= .07 (p= .035) Both the gratitude and
blessing intervention
decreased levels of
aggression. The
interventions did not dier
DeWall
(2012)
L 200 (T) College students 24 NR SGT (1-item: how
grateful they felt
that day)
Daily physical
aggression
. . . 25 days d=.42 (p< .01)* Controlling for positive
emotion, daily gratitude
predicted lower levels of daily
physical aggression.
DeWall
(2012)
L 168 (T) College students 32 NR SG (1-item: how
grateful they felt
during social
interaction)
Aggression in
response to
provocation
. . . 2 weeks Before controlling for
happiness d=1.08
(p< .001)*
After controlling not
able to calculate
Controlling for happiness felt
during interactions,
gratitude felt during
interactions was negatively
related to the percentage of
interactions where feelings
were hurt and how much
people expressed anger
outwardly toward the
person inicting hurt
DeWall
(2012)
E 79 (EG)
79 (ACG)
College students 33 NR . . . Behavioural
aggression
Gratitude letter (1x);
Provocation
manipulation
. . . Not able to calculate Provocation increased
aggression in the control
condition, it did not increase
aggression among grateful
participants; Among
insulted participants,
grateful participants
behaved less aggressively
than did control
participants; Among
participants who
experienced praise,
gratitude had no eect on
aggression
(Continued)
THE JOURNAL OF POSITIVE PSYCHOLOGY 7
Table 2. (Continued).
First
author (year)
Study
design
N
(condition) Type participant % Male
Mean Age
(SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
DeWall
(2012)
L 202 (T) College students 23 NR TG (GQ6) Aggression . . . 3 weeks β=.35 (p< .001) Analyses showed that higher
Time 2 empathy had
a signicant indirect eect
on the relationship between
Time 1 gratitude and Time 2
physical aggression,
controlling for Time 1
physical aggression, Time 1
and Time 2 positive aect,
and Time 2 gratitude.
Disabato
(2017)
L 797 (T) Multicultural adults 17 39 (14.20) TG (GQ6) Depression . . . 6 months d=.70 (p< .001)*
d=.07 (n.s.)*
Positive life-events are
a mediator between
gratitude and depression at
3 months but not at 6
months.
Flinchbaugh
(2012)
E 29 (SMT)
33 (GRAT)
22 (COM)
33 (CG)
College students 59 22 (1.50) . . . Perceived stress Gratitude journal
(weekly)
12 weeks Not able to calculate None of the conditions showed
a signicant eect on
perceived stress.
Jackowska
(2016)
E 40 (EG)
41 (ACG)
38 (WLCG)
Young adults 0 26 (0.77)
27 (0.79)
26 (0.82)
. . . Anxiety; Depression Gratitude journal (3x
a week for 2 weeks)
4 weeks Not able to calculate At follow-up the levels of
anxiety and depression
declined.
Jans-Beken
(2017)
L 706 (T) Adults 31 44 (14) TG (SGRAT) Psychopathological
symptoms
. . . 4 measures
during
7.5
months
β=.035 (n.s.) Trait gratitude is no predictor
on the long run for less
psychopathological
symptoms, accounting for
previous levels of
psychopathological
symptoms and subjective
well-being.
Jung (2017) E 17 (EG)
15 (CG)
Patients with
schizophrenia
NR NR TG (GRAT) Depression Gratitude disposition
promoting program
Twice
a week
for 4
weeks
η
2
= .11 (n.s.)* The results of the program did
not dier between the
experimental and the
control group.
Kerr (2015) E 16 (EG)
16 (ACG)
15 (CG)
Adults seeking
psychological
treatment
25 43 (11.1) SG (GAC) Psychological
functioning;
Depression,
anxiety and
stress
Gratitude journal (daily) 2 weeks η
2
= .37 (p< .001)
Not able to calculate
η
2
= .14 (p< .05)
η
2
= .11 (p< .05)
The gratitude intervention was
eective on psychological
functioning; No eect on
depression but there were
decreases on anxiety and
stress.
Khanna
(2016)
E 177 (T)
95 (EG)
82 (CG)
Highschool students 58 12 (0.67) . . . Negative
experience;
Negative aect
5 weekly sessions in
a classroom and
journal-based
homework
5 weeks Not able to calculate After the intervention the
levels of negative aect, and
negative experiences did
not change.
Killen (2015) E 88 (EG) Elderly 26 71 (7.51) TG (GQ6) Perceived stress TGT (daily for 2 weeks) 6 weeks R
2
= .05 (p< .01) Perceived stress declined over
the course of six weeks.
(Continued)
8L. JANS-BEKEN ET AL.
Table 2. (Continued).
First
author (year)
Study
design
N
(condition) Type participant % Male
Mean Age
(SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
Kleiman
(2013)
L 209 (T) College students 16 21 (4.12) TG (GQ6) Suicidal ideation;
Depressive
symptoms
. . . 4 weeks d=.44 (p< .01)* High levels of grit and
gratitude reduce suicide
ideation over time, with
gratitude as most important
predictor. Gratitude alone
nor grit alone were
associated with lower
suicide ideation.
Krentzman
(2015)
E 11 (EG)
12 (ACG)
Adults with
substance use
problems
52 46 (10.9) TG (GQ6) Negative aect TGT (daily for 2 weeks) 12 weeks d=.99 (p< .05)* The gratitude intervention
decreased the level of
negative aect over the
course of 12 weeks.
Lies (2014) L 310 (T) Earthquake
survivors
58 36 (10.5) TG (GQ6) Global distress;
PTSD
. . . 5 and 8
months
after
disaster
β=.051 (n.s.)
β= .052 (n.s.)
β= .061 (n.s.)
β= .052 (n.s.)
β=.008 (n.s.)
β=.232 (p< .001)
β=.207 (p< .001)
Gratitude at ve months after
the disaster did not predict
global distress or PTSD at
ve or eight months after
the disaster; Gratitude eight
months after the disaster
predicted global distress
and PTSD at eight months.
Martínez-
Martí
(2010)
E 41 (EG)
34 (ACG)
30 (ACG)
College students 11 21 (1.48) TG (GQ6); SG (GAC); Negative aect Gratitude journal (daily) 2 weeks Not able to calculate The gratitude intervention did
not have an eect on
negative aect.
Millstein
(2016)
L 156 (T) Older adults after
acute coronary
syndrome (ACS)
84 62 (10.6) TG (GQ6) Depression; Anxiety . . . 2 weeks
post-ACS
and 6
month
follow-up
β=.10 (p< .05)
β=.10 (p< .05)
There seems to be a negative
association between state
gratitude, and depression
and anxiety over the course
of 6 months, both outcomes
adjusted for baseline values,
gender, age, race, medical
and social risk factors, and
anxiety and depression.
OConnell
(2017)
E 63 (EG)
68 (ACG)
61 (CG)
Mainly young adult
sample
33 27 (12.63) TG (GQ6) Negative aect;
Depression
Reective behaviour
reective only
control journaling
3 times
a week
for 3
weeks
with 1
and 3
month
follow-up
η
2
= .07 (p< .05)
η
2
= .06 (p< .05)
Negative aect decreased in all
conditions after 1 month
with larger eects in the
reective behaviour
journaling but this decrease
disappeared at 3 months.
Depression decreased at
post-test but not at 1
month.
OLeary
(2015)
E 29 (EG)
22 (ACG)
10 (ACG)
Healthy adults 0 28 (6.65) . . . Perceived stress;
Depression
Gratitude journal
(4x a week)
3 weeks η
2
= .08 (n.s.)
η
2
= .09 (n.s.)
The gratitude intervention did
not signicantly lower levels
of stress or depression.
(Continued)
THE JOURNAL OF POSITIVE PSYCHOLOGY 9
Table 2. (Continued).
First
author (year)
Study
design
N
(condition) Type participant % Male
Mean Age
(SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
Otto (2016) E 34 (EG)
33 (ACG)
Women with breast
cancer diagnosis
057
(10.20)
SG (GAC) Fear of recurrence;
Death worry
Gratitude letter (once
a week for 6 weeks)
4.5 months Not able to calculate
d= .45 (p< .05)
The gratitude intervention did
not predict changes in fear
of recurrence but it was
a negative predictor for
death worry. This
association was mediated by
meaningful goal pursuit.
Ramírez
(2014)
E 26 (EG)
20 (CG)
Elderly 65 71
(7.06)
. . . Anxiety; Depression Gratitude letter 9 weeks η
2
= .21 (p< .001)
η
2
= .10 (p< .05)
The complete program reduces
anxiety and depression.
Salces-
Cubero
(2018)
E 36 (EG)
28 (ACG1)
28 (ACG2)
32 (CG)
Elderly 40 69 (7.78) . . . Anxiety;
Depression;
Negative aect
One-time activity with
gratitude, optimism
or savouring as key
variable
1 month η
2
= .43 (p< .001)
η
2
= .11 (p< .001)
No eect was found for
anxiety. For depression and
negative aect both the
between and within subject
models showed signicance.
Sirois (2017) L 163 (AR)
144 (IBD)
2 samples of
individuals with
arthritis (AR) and
irritable bowel
disease (IBD)
8 (AR)
21 (IBD)
AR: 47
(11.50)
IBD: 38
(13.00)
TG (GQ6) Depressive
symptoms;
. . . 6 months β=.22 (p< .01)
β=.14 (p< .05)
Gratitude was negatively
associated with depressive
symptoms over the course
of 6 months in patient with
arthritis and IBD, even when
adjusted for self-rated
health, pain, perceived
stress, social support, illness
cognitions, and
psychological thriving.
Southwell
(2017)
E 75 (EG)
52 (CG)
Adults with
depression and/
or anxiety
12 34 (10.80) TG (GQ6) Anxiety; Depression Daily diary for at least 3
x per week for 3
weeks
6 weeks η
2
= .26 (p< .001)
η
2
= .15 (p= .033)
η
2
= .38 (p< .01)
η
2
= .19 (p< .01)
η
2
=.04(p= .284)
η
2
= .03 (p= .371)
η
2
= .28 (p< .001)
η
2
= .04 (p= .318)
η
2
= .21 (p< .01)
Anxiety decreased pre- to post
test, post-test and follow-
up, and pre-test and follow-
up. Depression only
decreased between pre-test
and post-test, not post-test
and follow-up or pre-test
and follow-up. Stress
decreased pre-test to post-
test and pre-test t0 follow-
up, not between post-test
and follow-up.
Toepfer
(2012)
E 219 (T)
141 (EG)
78 (CG)
Adults 14 26 (11.00) TG (GQ6) Depression Gratitude letters (3
times)
4 weeks η
2
= .08 (p< .05)* The intervention decreased
depression.
Watkins
(2015)
E 47 (EG)
42 (ACG)
40 (CG)
College students 29 NR TG (SGRAT) Depression Gratitude journal (daily
for one week)
6 weeks η
2
= .06 (p< .01) The gratitude intervention
decreased depressive
symptoms over the course
of ve weeks after the
intervention.
(Continued)
10 L. JANS-BEKEN ET AL.
Table 2. (Continued).
First
author (year)
Study
design
N
(condition) Type participant % Male
Mean Age
(SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
Wolfe (2017) E 35 (EG)
28 (ACG)
45 (CG)
Graduate students 0 20 (6.93) . . . Body satisfaction;
Eating disorder;
Depression;
Negative aect
Gratitude listing daily
for 2 weeks
2 weeks η
2
= .14 (p< .001)*
η
2
= .09 (p< .01)*
η
2
= .01 (n.s.)*
η
2
= .08 (p< .05)*
η
2
= .01 (n.s.)*
η
2
= .12 (p< .01)*
η
2
= .08 (p< .05)*
The gratitude condition yielded
mix ndings on body
satisfaction and eating
disorder scales. The
gratitude condition
decreased depressive
symptoms and negative
aect.
Wong (2016) E 58 (EG)
56 (ACG)
53 (TAU)
Young adults in
psychotherapy
treatment
34 22 (5.00) . . . Psychological
symptoms such
as depression
and anxiety
Gratitude letter (3 in 3
weeks)
12 weeks d= .30 (p< .05) A combination of
psychotherapy and writing
gratitude letters improved
the mental health over the
course of 15 weeks.
Wong (2017) E 20 (EG) Graduate students 30 23 (3.76) SG (GAC) Psychological
distress
Gratitude Group
Program
2 months d= 1.19 (p< .001)
d= 1.37 (p< .001)
Participating in the Gratitude
Group Program reduced
psychological distress in
time between T1 and T2,
and T2 and T3.
Yang (2018) E (EG1)
(EG2)
(AEG)
Prisoners 100 35 (9.76) . . . Negative aect Counting blessings or
Random acts or
kindness for 6 weeks
6 weeks η
2
= .15 (p< .001)* Both the kindness and
gratitude interventions
decreased negative aect
with similar eect.
Note. L = longitudinal; E = experimental; T = total group, EG = experimental group; CG = control group; ACG = active control group; NR = not reported; TG = trait gratitude; SG = state gratitude; GQ6 = Gratitude
Questionnaire 6; GAC = Gratitude Adjectives Checklist; SGRAT = short gratitude, resentment and appreciation test; TGT = Three Good Things intervention; PTSD = post-traumatic stress disorder, ES = eect size, * =
estimation based on results in article.
THE JOURNAL OF POSITIVE PSYCHOLOGY 11
a longitudinal intervention study without control group,
showing small reductions in depressive symptoms
(Toepfer, Cichy, & Peters, 2012). The study of Southwell
and Gould (2017) showed lower levels of depression in
a clinical sample after a daily gratitude diary intervention.
Not all gratitude intervention studies in healthy adults
have yielded eects on psychopathology. No signicant
improvements in feelings of depression compared to
active control and wait-list conditions were shown in
a small-scale RCT study by OLeary and Dockray (2015).
No eects of gratitude letter writing on depression or
sadness were found in a sample with chronic back pain
(Baxter et al., 2012). A quasi-experimental study using
a gratitude disposition promotion program by Jung and
Han (2017) in patients with schizophrenia showed no
decrease in depressive symptoms after 4 weeks.
Several prospective observational studies have consis-
tently shown high levels of trait gratitude to be associated
with lower levels of depression, over periods up to 6
months, both in clinical and non-clinical samples with
small to moderate eect sizes (Disabato, Kashdan, Short,
&Jarden,2017; Millstein et al., 2016;Sirois&Wood,2017).
The negative association between trait gratitude and
depressive symptoms in the study of Disabato et al.
(2017) was partly explained by the experience of positive
life events and meaning in life, leading the researchers to
argue that gratitude as a personality strength may help to
motivate individuals with depression towards approach
behaviours, such as grateful acts, necessary to generate
positive life events and meaning in life, such as building
emotional intimacy with others.
Anxiety
RCTs targeting anxiety showed similar mixed results;
anxiety decreased over the course of the treatment
period (Jackowska et al., 2016; Ramírez et al., 2014;
Southwell & Gould, 2017; Wong, McKean Blackwell,
Goodrich Mitts, Gabana, & Li, 2017) and after 3 months
(Ramírez et al., 2014) in both healthy and clinical sam-
ples. On the other hand, the experimental study of
Salces-Cubero et al. (2018) could not provide evidence
for reduced anxiety after a gratitude training interven-
tion in elderly. Also, in a sample of adults with chronic
back pain, gratitude letter writing did not improve
anxiety (Baxter et al., 2012). One prospective observa-
tional studies have consistently shown high levels of
trait gratitude to be associated with lower levels of
anxiety, over periods up to 6 months, in a clinical sam-
ple with small eect sizes (Millstein et al., 2016).
Stress
RCTs looking into gratitude interventions reducing stress,
showed to decrease stress at the end of the intervention
period (Cheng et al., 2015; Southwell & Gould, 2017)and
after 3 months (Cheng et al., 2015). Similar results in lower
levels of perceived stress after a gratitude intervention
were found in a longitudinal study without a control
group in an elderly sample (Killen & Macaskill, 2015). The
small-scale RCT study by OLeary and Dockray (2015)
showed no reduced levels of stress after a gratitude jour-
naling intervention. Similarly, a study assigning under-
graduate students non-randomly to a weekly gratitude
journaling intervention, a stress management interven-
tion, a combination of the two, or a control condition,
did not nd any of the conditions to have a signicant
stress-reducing eect (Flinchbaugh et al., 2012).
Negative aect
The RCTs that included negative aect, all showed
a decrease of negative aect (OConnell et al., 2017b;
Salces-Cubero et al., 2018; Wolfe & Patterson, 2017;Yang,
Zhao, Aidi, & Kou, 2018). A mixed-methods randomized
controlled pilot among individuals in outpatient treatment
for alcohol use disorder showed the Three Good Things
exercise to moderately reduce negative aect compared to
placebo conditions (Krentzman et al., 2015). Martínez-Martí,
Avia, and Hernández-Lloreda (2010)didnotobserveany
changes in negative aect due to a two-week gratitude
journaling, any event journaling, or a hassles journaling
intervention in a small group of female participants. The
studybyChan(2011), although showing reduced emo-
tional exhaustion after count-your-blessings journaling,
did not reveal any changes in negative aect among parti-
cipants. Similar ndings,i.e.noeects of gratitude or acts
of kindness on negative emotions, were found by
Ouweneel, Le Blanc, and Schaufeli (2014). Two-week grati-
tude or kindness journaling in a small randomized study
among individuals on a waiting list for psychological treat-
ment did not reduce levels of negative aect, although
levels of anxiety were marginally reduced compared to the
control group (Kerr et al., 2015). Quasi-experimentally
designed gratitude drawing and educational interventions
in young children (Owens & Patterson, 2013)andadoles-
cents (Khanna & Singh, 2016), respectively, did not reduce
negative aectivity or negative experiences compared to
control conditions. The concept of gratitude may, however,
be dicult to grasp for children, especially when the chil-
dren are very young.
Other psychopathological distress
An RCT of WoWong et al. (2017)inasampleof
Chinese prisoners showed a decrease in psychologi-
cal distress directly after a social gratitude program
and after one-month follow-up. Emotional exhaus-
tion as symptoms of burnout decreased after a pre-
test/post-test design study, applying a gratitude
12 L. JANS-BEKEN ET AL.
intervention in a sample of Chinese school teachers.
These decreased levels of emotional exhaustion
were stronger in teachers with a high level of experi-
encing a meaningful life (Chan, 2011). An RCT by
Wong et al. (2016)showed,inasampleofyoung
adults seeking psychological counselling, that
a combination of psychotherapy and writing grati-
tude letters led to a larger improvement in global
mental health than psychotherapy alone or
a combination of psychotherapy and expressive writ-
ing. An RCT in women with early stage breast cancer
revealed that a weekly gratitude letter writing exer-
cise for 6 weeks did not induce changes in fear of
recurrence of the breast cancer, but levels of death
worry marginally decreased at 3 months after treat-
ment compared to the control condition (Otto,
Szczesny, Soriano, Laurenceau, & Siegel, 2016).
Kleiman, Adams, Kashdan, and Riskind (2013)
showed high levels of gratitude in synergy with
high levels of grit (i.e. perseverance and passion for
long-term goals; Duckworth, Peterson, Matthews, &
Kelly, 2007), to predict low levels of suicide ideation.
Gratitude levels at 5 months after exposure to
trauma, however, did not predict global distress
nor PTSD symptoms at ve or 8 months, according
to research by Lies, Mellor, and Hong (2014)among
earthquake survivors. The longitudinal study by
Jans-Beken, Lataster, Peels, Lechner, and Jacob
(2017) showed no prospective association between
trait gratitude and symptoms of psychopathology,
when taking into account previous levels of psycho-
pathology and subjective well-being.
Aggression
DeWall et al. (2012)conductedve studies with dierent,
large, mainly female samples of undergraduate students
and a variety of research designs to provide insight into
the relationship between aggression and gratitude. The
rst daily retrospective survey study (3 times a week for
a total of 25 days) showed that gratitude was negatively
associated with physical aggression, independent of the
level of positive emotions. A second two-week event
sampling study showed that feeling grateful seemed to
protect against hurt feelings and aggressive reactions due
to provocation within social interaction. In the third,
experimental study, participants were rst asked to write
an essay and a letter about ve things they were grateful
for or, for theactive controlgroup, about what they would
like to do. The participants then received, by manipula-
tion, either insulting or positive feedback on their essays
after which the researchers asked them to compete in
a reaction time task against the person who gave them
feedback. If the participants won, they could inict a blast
of white noise to the loser, which served as a measure of
aggression. While the participants who wrote a letter
about what they wanted to do showed signicantly
more aggression in the insult condition, participants
who wrote a gratitude letter did not show more aggres-
sion when provoked by insult. Lastly, DeWall et al. (2012)
investigated whether empathy mediates the negative
association between gratitude and aggression and dis-
covered that grateful individuals are in part less aggres-
sive because of their higher empathy for others. Deng
et al. (2018) found similar, small, eects regarding grati-
tude and aggression, but for both the gratitude and the
blessings condition. Chinese prisoners showed decreased
aggression after a group-based intervention, but the con-
tent of the intervention did not matter.
In conclusion
Findings from the longitudinal observational studies
included in this review are generally in line with ndings
from the considerable body of previous, largely cross-
sectional studies, suggesting negative associations
between trait gratitude and indicators of psychopathology
(Wood et al., 2010). High levels of trait gratitude thus seem
predictive of fewer symptoms of psychopathology in the
future, and approach behaviour motivation, contributing
to positive life events and meaning in life, may be impor-
tant mechanisms involved (Disabato et al., 2017).
Experimental studies, on the other hand, show very
mixed ndings regarding the eects of gratitude interven-
tions on indicators of psychopathology. Work by DeWall
et al. (2012) and Deng et al. (2018)further establishes gra-
titude as a social and moral emotion and being able to
empathize with others may prevent grateful individuals
from acting in an aggressive way (García-Sancho,
Salguero, & Fernández-Berrocal, 2014).
Psychological well-being
The rst component of well-being refers to autonomy and
personal mastery, a sense of purpose and meaning in life,
and personal growth and development (Ry,2014). Our
search yielded eight experimental, and four prospective
observational studies on the relationship between grati-
tude and psychological well-being, tapping into the
domains of (i) meaning in life, (ii) academic engagement,
(iii) basic psychological needs, (iv) self-esteem, (v) opti-
mism, (vi) humility, (vii) post-traumatic growth, (viii) resi-
lience, and (ix) dispositional hope. Table 3 represents an
overview of included studies.
Meaning in life
The quasi-experimental study by Flinchbaugh et al.
(2012) showed that, compared to a passive control
THE JOURNAL OF POSITIVE PSYCHOLOGY 13
Table 3. Psychological well-being: summary of articles.
First
author (year)
Study
design N (condition) Type participant
%
Male
Mean Age
(SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
Flinchbaugh
(2012)
E 29 (SMT)
33 (GRAT)
22 (COM)
33 (CG)
College students 59 22 (1.50) . . . Meaning in life;
Engagement
Gratitude journal
(weekly)
12 weeks η
2
= .08 (p= .03)
η
2
= .07 (p= .04)
A combination of stress
management and
gratitude journaling and
gratitude journaling
alone improved levels of
meaningfulness and
engagement.
Jackowska
(2016)
E 40 (EG)
41 (ACG)
38 (WLCG)
Young adults 0 26 (0.77)
27 (0.79)
26 (0.82)
. . . Optimism Gratitude journal
(3x a week for 2
weeks)
4 weeks Not able to calculate Optimism increased after
gratitude journaling.
Kerr (2015) E 16 (EG)
16 (ACG)
15 (CG)
Adults seeking
psychological
treatment
25 43 (11.10) SG (GAC) Optimism Gratitude journal
(daily)
2 weeks d= .97 (n.s.) No signicant increase in
optimism.
Kleiman
(2013)
L 209 (T) College students 16 21 (4.12) TG (GQ6) Meaning in life . . . 4 weeks β=.014 (p< .05)
β= .012 (p< .05)
Gratitude and the
synergistic eect of grit
and gratitude indirectly
predicted suicide
ideation through
changes in meaning in
life
Kruse (2014) L 25 (EG)
25 (ACG)
Undergraduate
students
NR 20 (1.83) TG (GQ6) Humility Daily online
questionnaires
for 2 weeks
γ
02
= .103, p< .05 Level of trait gratitude on
one day predicted
higher levels of humility
on the next day,
accounting for humility
on the previous day.
Lee (2015) L 127 (T) College students 27 20 (1.55) TG (Izards
Dierential
Emotion
Scale)
Basic psychological
needs
. . . 2 months R
2
= .12 (p<.01)
R
2
= .13 (p<.01)
R
2
= .09 (n.s.)
Gratitude showed an
upward spiral with
relatedness and
autonomy but not with
competence; gratitude
did not predict
competence.
Ouweneel
(2014)
E 25 (EG)
25 (CG)
College students 28 21 (1.93) . . . Academic
engagement
Gratitude journal
(daily)
5 days η
2
= .04 (n.s.)* Academic engagement did
not improve over time.
Peters (2013) E 26 (EG)
28 (ACG)
28 (CG)
Healthy adults 16 23 (11.75) . . . Dispositional
optimism;
Attributional
optimism
Imagery exercises
(daily)
1 week η
2
= .07 (n.s.)
Not able to calculate
The gratitude intervention
did not seem to be able
to increase both
dispositional as well as
attributional optimism.
Salces-Cubero
(2018)
E 36 (EG)
28 (ACG1)
28 (ACG2)
32 (CG)
Elderly 40 69 (7.78) . . . Resilience One-time activity
with gratitude,
optimism or
savouring as key
variable
1 month η
2
= .11 (p< .001)
η
2
= .17 (p< .001)
Gratitude and savouring
were eective both
between and within
subject, optimism was
not. This eect
disappeared after one
month.
(Continued)
14 L. JANS-BEKEN ET AL.
Table 3. (Continued).
First
author (year)
Study
design N (condition) Type participant
%
Male
Mean Age
(SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
Schnitker
(2018)
E 33 (normal)
37 (social)
31 (prayer)
Graduate students 12 18 (0.42) TG (GQ6) Dispositional hope Dierent kinds of
gratitude
journaling once
a week for 5
weeks
6 weeks Not able to calculate The prayer condition in
participants with high
eort improved
dispositional hope.
Zhou (2015) L 217 (T) Adolescent
earthquake
survivors
50 14
(1.39)
TG (GQ6) Post-traumatic
growth
. . . 3.5, 4.5, and 5.5 year
after the disaster
Not able to calculate Gratitude 3.5 and 4.5 years
after the event
predicted post-
traumatic growth at 4.5
and 5.5 years after the
event; Gratitude 3.5
years after the event
predicted post-
traumatic growth at 5.5
years after the event
through deliberate
rumination at 4.5 years
after the event.
Wong (2017 E 20 (EG) Graduate students 30 23 (3.76) SG (GAC) Meaning in life Gratitude Group
Program
2 months d=2.98 (p< .01)
d=2.77 (p< .01)
Participating in the
Gratitude Group
Program increased the
level of meaning in life
in time between T1 and
T2, and T2 and T3.
Note. L = longitudinal; E = experimental; T = total group, EG = experimental group; CG = control group; ACG = active control group; TG = trait gratitude; SG = state gratitude; GQ6 = Gratitude Questionnaire 6; GAC =
Gratitude Adjectives Checklist; SMT = stress management techniques; GRAT = gratitude journaling; COM = combination SMT and GRAT, ES = eect size, * = estimation based on results in article.
THE JOURNAL OF POSITIVE PSYCHOLOGY 15
condition, gratitude journaling increased the level of
meaningfulness in undergraduate students over the
course of 12 weeks, a small eect that was amplied
when adding a stress management training. The study
of Otto et al. (2016) showed that a 6-week online
gratitude intervention increased meaningful goal pur-
suit in a sample of women with breast cancer compared
to the control group. Also, WoWong et al. (2017) found
that meaning in life increased after a gratitude group
intervention over the course of 2 months, although in
a small study (N = 20) with nonrandomized, one-group
pre-post-test design. In line with these ndings,
Kleiman et al. (2013) observed without intervening
that trait gratitude and grit work synergistically in pro-
tecting against suicidal ideation through increased
meaning in life.
Academic engagement
In addition to its small eects on meaning in life in the
quasi-experimental study by Flinchbaugh et al. (2012),
gratitude journaling increased course engagement
(purpose in life) in undergraduates, and this small eect
was further amplied by additionally providing stress
management strategies. Ouweneel et al. (2014), how-
ever, showed no benecial eect of a gratitude inter-
vention on academic engagement when compared to
a kindness or neutral intervention in a group of under-
graduate students in their RCT study.
Basic psychological needs
A prospective study by Lee, Tong, and Sim (2015)
revealed associations between gratitude and psychologi-
cal need fullment: gratitude predicted relatedness and
autonomy, although not competence, over time. In line
with these ndings, Kerr et al. (2015) found a gratitude
intervention to improve feelings of relatedness in
a clinical sample awaiting psychological treatment.
Self-esteem
A positive eect of gratitude interventions on self-
esteem has been suggested by ndings from an experi-
ment by Rash et al. (2011), in which participants were
randomly instructed to recall either grateful feelings for
someone or something, or a memorable event twice
a week for a total of 4 weeks. After 4 weeks, participants
in the gratitude condition showed higher self-esteem
than participants in the control condition. No increase
in self-esteem was observed, however, in young chil-
dren participating in the gratitude drawing intervention
study by Owens and Patterson (2013).
Optimism
Three RCTs have investigated the eects of gratitude
interventions on feelings of optimism. A 5-min daily
imagery and writing intervention in a healthy study
sample did not elicit changes in optimism (Peters
et al., 2013), but gratitude journaling, on the other
hand, resulted in increased optimism compared to
active control and no-treatment groups in the studies
by Kerr et al. (2015) and Jackowska et al. (2016).
Humility
An observational study by Kruse, Chancellor, Ruberton,
and Lyubomirsky (2014), asking participants to ll in
daily questionnaires during a period of 2 weeks,
showed that trait gratitude at the previous measure-
ment was weakly and positively associated with humi-
lity at the next measurement.
Post-traumatic growth
One study on post-traumatic growth in adolescent sur-
vivors of the Sichuan earthquake, Zhou and Wu (2015)
observed that gratitude at 3.5 and 4.5 years after the
event predicted post-traumatic growth at 4.5 and 5.5
years, and that this association was at least partly
mediated by the process of deliberate rumination.
Resilience
One quasi-experimental study showed that, compared
to a passive control condition, both a gratitude and
a savouring intervention increased resilience with
a small eect in an elderly sample over the course of
1 month, whereas an optimism intervention did not
(Salces-Cubero et al., 2018).
Dispositional hope
Schnitker and Richardson (2018) conducted a study
with three conditions. Participants had to write down
10 things they were grateful for over the past week. The
rst group had to read the list aloud to themselves,
the second group had to read the list aloud to a friend
or signicant other, and the third group had to read the
list aloud to God. The researchers framed this last con-
dition as prayer and participants in this group showed
an increase in dispositional hope, compared to the
other two conditions, especially when the participants
executed the exercises with high eort. However, parti-
cipants were students from a private Christian univer-
sity, and thus it cannot be assumed that eects will
generalize to non-Christian samples.
In conclusion
Although results on gratitude and factors of psychological
well-being are currently scant, fragmented, and
16 L. JANS-BEKEN ET AL.
inconclusive, there is some evidence suggesting that gra-
titude, possibly in synergy with other psychological com-
petencies, may contribute to enhanced psychological
well-being. However, reported eects are small which
warrants further examination of the benet of state and
trait gratitude as a predictor for psychological well-being.
Emotional well-being
The second component of mental well-being is emo-
tional well-being, comprising components such as hap-
piness and life satisfaction (Diener, 2000). Our search
yielded 32 intervention studies, and 1 prospective
observational study on the relationship between grati-
tude and (i) happiness, (ii) subjective well-being, (iii)
positive aect, (iv) life satisfaction, (v) ourishing, and
(vi) quality of life. Table 4 presents an overview of the
included articles.
Happiness
Expressing gratitude in an Islamic and secular version of
an intervention showed to be associated with happi-
ness during an RCT of 16 weeks of Al-Seheel and Noor
(2016), compared to the control group. Another RCT
during 5 weeks in a sample of healthy adult women
improved happiness after a gratitude intervention, but
a mindfulness intervention seemed more eective; the
control group did not improve during this period
(OLeary & Dockray, 2015). A more recent quasi-
experimental study of Salces-Cubero et al. (2018)inan
elderly sample, resulted in increased happiness in both
the gratitude and the savouring condition, but not the
optimism condition, compared to a passive control con-
dition, for the between-subjects model and the within-
subjects model over the course of a month. The pilot
study of Baxter et al. (2012) showed an increase in daily
happiness in a small sample of respondents with
chronic back pain over the course of ve to 7 weeks.
The positive psychology intervention of Ramírez et al.
(2014) generated an increase in happiness in elder
participants at post-intervention, but levels of happi-
ness returned to base levels after 4 months after the
intervention.
Subjective well-being
Watkins et al. (2015) performed an RCT in an undergradu-
ate student sample during 5 weeks. Subjective well-being
continued to increase, also after the intervention had
ended after 1 week, compared to a memory or placebo
intervention. Also, in a pre-post-follow-up trial, subjective
well-being improved with small to moderate eect in the
gratitude journaling intervention conducted with a sample
of distressed participants (Southwell & Gould, 2017).
Similarly, WoWong et al. (2017) found that subjective well-
being increased after a gratitude group intervention imme-
diately after the intervention but not at follow-up.
A nonrandomized one-group pre-post-test design in
Chinese prisoners showed that both a gratitude sharing
intervention and a counting blessing intervention
increased subjective well-being of the inmates over the
course of 5 weeks (Deng et al., 2018). Yang et al. (2018)
also found eects for both the experimental and active
control condition: similar to the gratitude intervention,
the kindness intervention increased the reported levels of
subjective well-being in a sample of male prisoners with
small to moderate eect. The Mobile delivery of gratitude
interventions to increase subjective well-being was
assessed in two small-scale pilot-RCTs by Ghandeharioun,
Azaria, Taylor, and Picard (2016) with promising results: use
of their Kind and Gratefulapp led to increased practice of
gratitude, increased positive emotional valence and
decreased emotional arousal, and increased levels of sub-
jective well-being compared to baseline. Layous, Nelson,
Kurtz, and Lyubomirsky (2017) conducted two RCTs. In the
rst study, they did not nd an increase in subjective well-
being by any of the conditions: general gratitude, specic
gratitude, optimism, joy or control. In their second study,
again they did not nd a direct eect of conditions on
subjective well-being, but they did nd that the conditions
as mentioned above increased elevation, which in turn
increased subjective well-being (Layous et al., 2017). One
prospective study by Jans-Beken et al. (2017)showed
a small positive association between trait gratitude and
levels of subjective well-being, when accounting for pre-
vious levels of psychopathology and subjective well-being.
Positive aect
Writing gratitude letters increased positive aect with
every letter written during a period of 3 weeks (Toepfer
et al., 2012). Positive aect increased in an RCT with
a gratitude and active control intervention, and
a control group (Ouweneel et al., 2014; Wolfe &
Patterson, 2017). However, positive aect decreased to
baseline level 2 weeks after the intervention while par-
ticipants from the kindness intervention group reported
higher positive aect at that point of time (Ouweneel
et al., 2014). In an RCT by Otto et al. (2016), participants
diagnosed with breast cancer showed a stable level of
positive aect during the study period of 4,5 months, in
contrast to the control group who showed a decline of
positive aect during the same period. A similar nding
is reported by OConnell et al. (2017b); positive aect
stayed fairly the same in the study but aect balance
improved after both a reective interpersonal gratitude
journaling (fostering gratitude) intervention and
areective behavioural interpersonal gratitude
THE JOURNAL OF POSITIVE PSYCHOLOGY 17
Table 4. Emotional well-being: summary of articles.
First
author (year)
Study
design N (condition) Type participant
%
Male Mean Age (SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
Al-Seheel
(2016)
E 20 (EG)
19 (ACG)
21 (CG)
Undergraduate
students
15 22 (1.22) . . . Happiness Gratitude journal
(daily for 2
weeks) +
gratitude letter
17 days η
2
= .28 (p< .05) The participants in the
Islamic condition
showed a higher
increase in happiness
but there was no
signicant dierences
between the dierent
conditions.
Baxter
(2012)
E 8 (T)
4(EG)
4(ACG)
Adults with chronic
back pain
50 55 (8.25) . . . Daily happiness Character strength
and gratitude
intervention
57 weeks Not able to calculate The Character strength
and gratitude
intervention increased
daily happiness.
Carson
(2010)
E 9(T) Service users
attending
a community
mental health
team
NR NR TG (GQ6) Being thankful; Life
satisfaction;
Environmental
mastery; Social
feelings
Two 2 hour
workshop with
daily dairies
1 month Not able to calculate After the intervention
participants reported
being thankful for
more things,
increased life
satisfaction,
environmental
mastery, and social
feelings.
Chan
(2010)
E 96(T) Chinese school
teachers
18 33 (7.57) TG (GQ6) Life satisfaction;
Positive aect; SG
(GAC)
Weekly log of TGT
and Naikan
questions
8 weeks d=.513 (p< .001)
d=.486 (p< .01)
d=.390 (p< .05)
The intervention
increased life
satisfaction in the
low-trait gratitude
group, and positive
aect
Chan
(2011)
E 63(T) Chinese school
teachers
16 34 (6.91) TG (GQ6) Life satisfaction Weekly log of TGT
and Naikan
questions
8 weeks d=.54 (p< .01) The intervention
increased life
satisfaction in the
high meaningful-life
group
Deng (2018) E 29 (EG)
37 (ACG)
30 (CG)
Male prisoners 100 35 (9.65) TG (GRAT) Subjective well-
being
5 daily diary
(counting
blessing) versus 5
weekly group
sessions (sharing
gratitude)
5 weeks η
2
= .11 (p= .005) Both the gratitude and
blessing intervention
increased levels of
SWB. The
interventions did not
dier
Flinchbaugh
(2012)
E 29 (SMT)
33 (GRAT)
22 (COM)
33 (CG)
College students 59 22 (1.50) . . . Life satisfaction Gratitude journal
(weekly)
12 weeks Not able to calculate None of the
interventions were
able to increase levels
of life satisfaction.
(Continued)
18 L. JANS-BEKEN ET AL.
Table 4. (Continued).
First
author (year)
Study
design N (condition) Type participant
%
Male Mean Age (SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
Ghandeharioun
(2016)
E 27 (T) Young adults 48 NR TG (GQ6) Thankful behaviour Kind and grateful
app use
5 weeks Not able to calculate Thankful behaviour
increased.
Psychological well-
being and trait
gratitude increased
during the 5 weeks.
The best moment for
expressing gratitude is
after social
interaction, physical
activity and location
change
Işık (2017) E 11 (EG)
10 (CG)
First year college
students
33 18 (2.47) TG (GQ6) Life satisfaction;
Positive aect
3 week daily
reective
gratitude
journaling
3 weeks Not able to calculate Both life satisfaction and
positive aect
increased, using
Wilcoxon Signed Rank
Tests
Jackowska
(2016)
E 40 (EG)
41 (ACG)
38 (WLCG)
Young adults 0 26 (0.77)
27 (0.79)
26 (0.82)
. . . Flourishing Gratitude journal
(3x a week for 2
weeks)
4 weeks Not able to calculate The intervention did not
increase levels of
ourishing.
Jans-Beken
(2017)
L 706 (T) Adults 31 44 (14) TG (SGRAT) Subjective well-
being
. . . 4 measures
during 7.5
months
β= .088 (p< .001) Trait gratitude is
a predictor on the
long run for increased
well-being,
accounting for
previous levels of
psychopathological
symptoms and
subjective well-being.
Jung (2017) E 17 (EG)
15 (CG)
Patients with
schizophrenia
NR NR TG (GRAT) Life satisfaction Gratitude
disposition
promoting
program
Twice a week
for 4 weeks
η
2
= .25 (p< .01)* The results of the
program diered
between the
experimental and the
control group; the
experimental group
reported higher life
satisfaction than the
control group.
(Continued)
THE JOURNAL OF POSITIVE PSYCHOLOGY 19
Table 4. (Continued).
First
author (year)
Study
design N (condition) Type participant
%
Male Mean Age (SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
Khanna
(2016)
E 177 (T)
95 (EG)
82 (CG)
Highschool students 58 12 (0.67) . . . Positive mental
health; Positive
experience; Life
satisfaction;
Social-cognitive
perceptions of
gratitude;
Positive aect; SG
(GAC)
5 weekly sessions in
a classroom and
journal-based
homework
5 weeks η
2
= .01 .10 (p< .05) After the intervention
the levels of
psychological well-
being, positive mental
health total score,
social-cognitive
perception of
gratitude, positive
aect, positive and
balanced experiences,
state gratitude, and
life satisfaction
increased but
disappeared after
controlling for scores
on Time 1
Killen
(2015)
E 88 (EG) Elderly 26 71 (7.51) TG (GQ6) Flourishing;
Emotional
balance
TGT
(daily for 2 weeks)
6 weeks η
2
= .10 (p< .001)
η
2
= .04 (n.s.)
Flourishing increased
across time, emotional
balance did not.
Layous (2017) E 45 (general gratitude)
47 (specic gratitude
48 (optimism
47 (joy)
46 (ACG)
Graduate students 30 20 (2.92) . . . Subjective well-
being
Random act of
kindness once
a week for 3
weeks
6 weeks There were no changes
over time in
subjective well-being
(T1-T2 and T2-T3) and
there were no group
dierences (T1-T2 and
T2-T3)
Layous (2017) 34 (EG1)
34 (EG2)
36 (EG3)
35 (ACG)
Graduate students 24 20 (3.10) . . . Subjective well-
being
Activity once a week
for 7 weeks
3 months Across conditions, well-
being did not increase
over time and no
dierences between
groups were detected.
OConnell
(2017a)
E 63 (EG)
68 (ACG)
61 (CG)
Mainly young adult
sample
33 27 (12.63) TG (GQ6) Aect balance; Life
satisfaction;
Positive aect
Reective
behaviour
reective only
control
journaling
3 times a week
for 3 weeks
with 1 and 3
month
follow-up
η
2
= .06 (p< .05) Reective behaviour
improved aect
balance at post-test.
There were no
dierences in life
satisfaction and
positive aect in the
dierent condition.
(Continued)
20 L. JANS-BEKEN ET AL.
Table 4. (Continued).
First
author (year)
Study
design N (condition) Type participant
%
Male Mean Age (SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
OConnell
(2017b)
E 29 (EG1)
29 (EG2)
30 (ACG)
Student population 42 24 (7.79) TG (GQ6) Life satisfaction Gratitude journaling
4 days in 2 weeks
8 weeks η
2
= .09 (p< .01) Life satisfaction was
higher in participants
in the traditional
gratitude journaling
condition compared
to the control group.
There was no eect
found for the
interpersonal
gratitude journaling
condition.
OLeary
(2015)
E 29 (EG)
22 (ACG)
10 (ACG)
Healthy adults 0 28 (6.65) . . . Happiness Gratitude journal
(4x a week)
3 weeks η
2
= .07 (n.s.) No signicant increase in
levels of happiness
after the intervention.
Otto
(2016)
E 34 (EG)
33 (ACG)
Women with breast
cancer diagnosis
0 57 (10.20) SG (GAC) Positive aect Gratitude letter
(once a week for
6 weeks)
4.5 month Not able to calculate The slope of the
gratitude intervention
remained stable
whereas the slope of
the control condition
declined signicantly.
The gratitude
intervention
prevented positive
aect from declining
over the course of 4.5
month.
Ouweneel
(2014)
E 25 (EG)
25 (CG)
College students 28 21 (1.93) . . . Academic
engagement
Gratitude journal
(daily)
5 days η
2
= .04 (n.s.)* Academic engagement
did not improve over
time.
Owens
(2013)
E 22 (EG)
23 (ACG)
17 (ACG)
Children 48 7 (1.73) . . . Positive aect; Life
satisfaction
Drawing (once
a week)
4 to 6 weeks Not able to calculate
η
2
= .06 (n.s.)*
None of the drawing
interventions
increased positive
aect or life
satisfaction.
Peters
(2013)
E 26 (EG)
28 (ACG)
28 (CG)
Healthy adults 16 23 (11.75) . . . Life satisfaction Imagery exercises
(daily)
1 week η
2
= .018 (p< .05)* The gratitude
intervention was able
to increase levels of
life satisfaction.
Proyer
(2013)
E 39 (EG)
44 (ACG)
53 (CG)
Healthy adults 41 41 (13.08) . . . Life satisfaction Gratitude letter and
4 other
interventions
NR η
2
= .08 (p< .05)* The program was able to
increase levels of life
satisfaction.
(Continued)
THE JOURNAL OF POSITIVE PSYCHOLOGY 21
Table 4. (Continued).
First
author (year)
Study
design N (condition) Type participant
%
Male Mean Age (SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
Rash
(2011)
E 56 (T)
NR (EG)
NR (ACG)
Adult
community
sample
54 23 (3.00) TG (GQ6); Life satisfaction Gratitude journal
(twice a week)
4 weeks η
2
= .10 (p< .05)
d= .63 (p< .05)*
The gratitude
intervention increased
the levels of life
satisfaction; Trait
gratitude moderated
the association
between the
intervention and life
satisfaction.
Ramírez (2014) E 26 (EG)
20 (CG)
Elderly 65 71
(7.06)
. . . Positive memories;
Life satisfaction;
Happiness
Gratitude letter 9 weeks η
2
= .15 (p< .001)
η
2
= .10 (p< .05)
η
2
= .09 (p< .05)
The program was able to
increase levels of
positive memories, life
satisfaction, and
happiness.
Salces-Cubero
(2018)
E 36 (EG)
28 (ACG1)
28 (ACG2)
32 (CG)
Elderly 40 69 (7.78) . . . Happiness; Positive
aect; Life
satisfaction
One-time activity
with gratitude,
optimism or
savouring as key
variable
1 month η
2
= .10 (p< .001)
η
2
= .09 (p< .001)
η
2
= .22 (p< .001)
η
2
= .32 (p< .001)
η
2
= .25 (p< .001)
η
2
= .23 (p< .001)
Gratitude and savouring
increased happiness,
positive aect, and life
satisfation, in both
between and within
subject models,
optimism was not.
This eect remained
the same after one
month.
Schnitker
(2018)
E 33 (normal)
37 (social)
31 (prayer)
Graduate students 12 18 (0.42) TG (GQ6) Life satisfaction Dierent kinds of
gratitude
journaling once
a week for 5
weeks
6 weeks Not able to calculate The prayer condition in
participants with high
eort improved life
satisfaction.
Southwell
(2017)
E 75 (EG)
52 (CG)
Adults with
depression and/
or anxiety
12 34 (10.80) TG (GQ6) Subjective well-
being
Daily diary for at
least 3 x per
week for 3 weeks
6 weeks η
2
= .27 (p< .001)
η
2
= .03 (p= .004)
Subjective well-being
increased pre- to
post-test and pre-test
and follow-up.
Toepfer(2012) E 219 (T)
141 (EG)
78 (CG)
Adults 14 26 (11.00) TG (GQ6) Life satisfaction;
Happiness
Gratitude letters (3
times)
4 weeks η
2
= .24 (p< .001)*
η
2
= .06 (p< .01)*
The intervention
increased life
satisfaction and
happiness.
Watkins (2015) E 47 (EG)
42 (ACG)
40 (CG)
College students 29 NR TG (SGRAT) Subjective well-
being
Gratitude journal
(daily for one
week)
6 weeks η
2
= .054 (p< .05) A signicant rise of
subjective well-being
was not apparent at
post-test and 1 week
follow-up but it was
on the 5 week follow-
up.
Wolfe (2017) E 35 (EG)
28 (ACG)
45 (CG)
Graduate students 0 20 (6.93) . . . Positive aect Gratitude listing
daily for 2 weeks
2 weeks η
2
= .06 (p< .05)* The gratitude condition
improved positive
aect
(Continued)
22 L. JANS-BEKEN ET AL.
Table 4. (Continued).
First
author (year)
Study
design N (condition) Type participant
%
Male Mean Age (SD)
Gratitude
measure(s)
Dependent
variable(s) Intervention Time frame ES Summary of ndings
Wong (2017) E 20 (EG) Graduate students 30 23 (3.76) SG (GAC) Subjective well-
being
Gratitude Group
Program
2 months d=2.92 (p< .01)
d=1.51 (n.s.)
Participating in the
Gratitude Group
Program increased the
level of subjective
well-being in time
between T1 and T2,
but not between T2
and T3.
Yang (2018) E (EG1)
(EG2)
(AEG)
Prisoners 100 35 (9.76) . . . Subjective well-
being; Positive
aect; Life
satisfaction;
Subjective vitality
Counting blessings
or Random acts
or kindness for 6
weeks
6 weeks η
2
= .14 (p< .001)*
η
2
= .29 (p< .001)*
η
2
= .26 (p< .001)*
η
2
= .08 (p< .001)*
Both the kindness and
gratitude
interventions
increased all
dependent variables.
Kindness scores better
with life satisfaction
and vitality, for
positive aect and
well-being both
interventions were
similarly eective.
Note. L = longitudinal; E = experimental; T = total group, EG = experimental group; CG = control group; ACG = active control group; NR = not reported; TG = trait gratitude; SG = state gratitude; GQ6 = Gratitude
Questionnaire 6; GAC = Gratitude Adjectives Checklist; SGRAT = short gratitude, resentment, and appreciation test; SMT = stress management techniques; GRAT = gratitude journaling; COM = combination SMT and
GRAT; TGT = Three Good Things intervention, ES = eect size, * = estimation based on results in article.
THE JOURNAL OF POSITIVE PSYCHOLOGY 23
journaling (fostering gratitude and expression of grati-
tude) intervention, compared to the control group. The
study of Işık and Ergüner-Tekinalp (2017) showed that
keeping a gratitude journal for 3 weeks, compared to
a passive control condition, increased the level of posi-
tive aect in rst-year college students. A quasi-
experimental study of Salces-Cubero et al. (2018)inan
elderly sample, resulted in increased positive aect in
both the gratitude and the savouring condition, not the
optimism condition, for the between-subjects model
and the within-subjects model over the course of
a month. Yang et al. (2018) also found results for both
the experimental and active control condition. Just like
the gratitude intervention, the kindness intervention
increased the reported levels of positive aect in
a sample of male prisoners with small to moderate
eect. Another quasi-experimental design in a sample
of middle school students showed higher positive aect
after the Frohs 5-week Gratitude Curriculum compared
to a control group (Khanna & Singh, 2016). The only
study that did not show increased positive aect was in
a sample of elementary school students; positive aect
did not increase after a drawing intervention (Owens &
Patterson, 2013).
Life satisfaction
Positive psychology interventions, which target grati-
tude among others, showed to increase levels of life
satisfaction in undergraduate student, adults and
elderly (Flinchbaugh et al., 2012; Proyer, Ruch, &
Buschor, 2013; Ramírez et al., 2014). Stand-alone grati-
tude interventions such as journaling, gratitude letters,
and Frohs Gratitude Curriculum showed to increase in
reported life satisfaction. (Carson, Muir, Clark, Wakely, &
Chander, 2010; Chan, 2011;Işık & Ergüner-Tekinalp,
2017; Khanna & Singh, 2016; Salces-Cubero et al.,
2018; Schnitker & Richardson, 2018; Toepfer et al.,
2012). Rash et al. (2011) observed that trait gratitude
moderated the eects of a gratitude intervention on
satisfaction with life, such that those with low trait
gratitude beneted from the intervention but those
high in trait gratitude not; a nding that was previously
reported by Chan (2011). Yang et al. (2018) also found
eects for both the experimental and active control
condition. Similar to the gratitude intervention, the
kindness intervention increased the reported levels of
life satisfaction in a sample of male prisoners with small
to moderate eect. However, not all studies yielded
positive eect between gratitude and life satisfaction.
Peters et al. (2013) observed no improvement of life
satisfaction following a one-week gratitude intervention
in adults. After a journaling intervention, life satisfaction
stayed about the same in a sample of both healthy
students and adults (OConnell et al., 2017b). Results
from another RCT (OConnell, OShea, and Gallagher
(2017a) showed that a traditional gratitude intervention
was eective in increasing life satisfaction over the
course of 8 week in a sample of students, compared
to the control group. The eect, however, was quite
small, and no dierences were found in this study for
the interpersonal gratitude intervention compared to
the control group and the traditional gratitude inter-
vention. In the drawing study with children by Owens
and Patterson (2013), just as with positive aect, life
satisfaction did not increase.
Flourishing
Flourishing was included in a study by Killen and
Macaskill (2015). They used the Three Good Things
intervention in a sample of sub-clinical elderly and
showed that ourishing increased during and after the
2-week intervention, up to 45 days after baseline mea-
sure. However, the gratitude journaling intervention of
Jackowska et al. (2016) was ineective on levels of
ourishing.
Quality of life
Jung and Han (2017) developed a gratitude disposition
promotion program that is intended for chronic schizo-
phrenic patients. The program consists of several
stages, associated to dierent components of trait gra-
titude: improvement of sensitivity to external stimuli,
understanding of gratitude, expression of gratitude,
and empathy related to gratitude. Except that the pro-
gram was able to indeed increase trait gratitude, it also
showed to improve reported quality of life within this
sample.
In conclusion
Taken together, although not all studies have yielded
positive results, the vast majority of research shows
measures of emotional well-being to increase with
small to moderate positive eects in response to
a variety of gratitude interventions administered in
a variety of populations, or to maintain at a certain
level and prevent emotion well-being from declining.
Social well-being
The third component of well-being is social well-being
which embodies social and societal participation, consist-
ing of social skills, social contacts and meaningful relation-
ships, and societal commitments and a purposeful
employment (Keyes, 1998). Our search identied 19
experimental, and 7 prospective studies investigating
the relationship between gratitude and social and societal
24 L. JANS-BEKEN ET AL.
Table 5. Social well-being: summary of articles.
First
author
(year)
Study
design N (condition) Type participant
%
Male
Mean Age
(SD) Gratitude measure(s) Dependent variable(s) Intervention Time frame ES Summary of ndings
Algoe
(2016)
E 24 couples (EG)
23 couples (ACG)
Couples in
a marriage,
engaged, or
exclusively
dating
50 29 (7.50) . . . Relationship satisfaction;
Satisfaction with life;
Positive emotions;
Negative emotions;
Daily ability to adapt;
Daily relationship
evaluation; Daily life
satisfaction
46 random signaled
conversations about
the things the partner
did and one is grateful
for
4 weeks Not able to calculate When adjusting for the partners
responsiveness the gratitude
intervention seemed to be
able to increase positive
emotions and the ability to
adapt, not the other outcome
measures. The authors think
this might be because the
couples scored very high on
measures of life satisfaction at
the start of the study.
Caleon
(2017)
E 46 (EG)
57 (CG)
College students 48 13 (NR)
15 (NR)
TG (GQ6) Relatedness with parents,
teachers, and peers
4dierent activities in
school
2 weeks η
2
= .04 (p= .03) The students who participated in
the intervention class only
reported improved relatedness
with parents and peers, not
with teachers.
Cho
(2012)
E 183 (T)
NR (low power)
NR (high power
College students 55 20 (NR) . . . Denigration Notes with or without
a gratitude expression
... η
2
= .07 (n.s.)* High-power individuals whose
competence was threatened
denigrated their subordinates.
This pattern disappeared
when the subordinate
expressed gratitude.
Among low-power participants,
there were no main eects of
competence and gratitude
expression, nor an interaction
between competence and
gratitude expression.
Cho
(2012)
E 123 (T)
NR (low power)
NR (high power)
College students 56 20 (NR) . . . Denigration Notes with or without
a gratitude expression
. . . Not able to calculate Gratitude expression ameliorates
aggressive tendencies of
threatened individuals with
high-power by increased
feelings of social worth in the
eyes of ones subordinates.
Converse
(2012)
E 42 pairs of
strangers
NR (ongoing)
NR (completed)
Adults NR NR SG (1 item asking for
gratitude for
helper)
Instrumentality of help Trivia quiz with helpline . . . Contestants in the ongoing-game
condition were more
appreciative for the helpline
than those in the completed
game condition who even
received more assistance.
(Continued)
THE JOURNAL OF POSITIVE PSYCHOLOGY 25
Table 5. (Continued).
First
author
(year)
Study
design N (condition) Type participant
%
Male
Mean Age
(SD) Gratitude measure(s) Dependent variable(s) Intervention Time frame ES Summary of ndings
Converse
(2012)
E 40 (T)
NR (active)
NR (completed)
College students
working in
active
participation
43 NR SG (1 item asking for
appreciation of
help
Instrumentality of help . . . 3 months d= 1.03 (p< .01) Appreciation was related to the
current level of
instrumentality; Beneciaries
appreciation of earlier help
was more a function of their
current reliance on their prior
helpers than a function of
their satisfaction with the
preceding semesters grade.
Converse
(2012)
E 114 (T) College students NR NR SG (1 item asking
form appreciation
of help)
Success of the goal . . . . . . d= 0.53 (p< .05) If assistance leads to success,
appreciation will decrease
when new goals take priority.
Diebel
(2016)
E 49 (EG)
51 (ACG)
Primary school
children
51 9 (NR) SG (adapted GQ6) Belonging at school Gratitude diary (5x a week
for 4 weeks)
4 weeks η
2
= .23 (p< .001) The gratitude diary was able to
increase the sense of
belonging at school in primary
school children.
Froh
(2010)
L 700 (T) Middle school
students
48 12 (0.89) SG (GAC) Social integration . . . 6 months d=.23 (p< .01)* Gratitude predicts social
integration through prosocial
behaviour.
Grant
(2010)
E 35 (EG)
34 (CG)
College students 36 22 (3.55) . . . Prosocial behavior E-mails with or without
a gratitude expression
. . . Not able to calculate Gratitude expressions increase
prosocial behavior through
enabling helpers to feel more
socially valued, rather than
through enabling helpers to
feel more ecacious or
through positive or negative
aect
Grant
(2010)
E 29 (EG)
28 (CG)
College students 49 23 (3.47) . . . Prosocial behavior E-mails with or without
a gratitude expression
. . . Not able to calculate Social worth, but not self-ecacy,
positive aect, negative aect,
or empathy, mediated the
eect of an expression of
gratitude from one beneciary
on prosocial behavior directed
toward a dierent beneciary.
Grant
(2010)
E 20 (EG)
21 (CG)
Fundraisers 24 NR . . . Prosocial behavior Expression of gratitude
from director
2 weeks Not able to calculate Expressing gratitude increased
the prosocial behavior.
Strengthened the fundraisers
feelings of social worth, not by
enhancing their feelings of
self-ecacy
Grant
(2010)
E 79 (T)
NR (EG)
NR (CG)
College students 32 NR . . . Prosocial behavior Expression of gratitude
directly
. . . Not able to calculate Social worth predicted higher
prosocial behavior in the
gratitude expression
condition.
(Continued)
26 L. JANS-BEKEN ET AL.
Table 5. (Continued).
First
author
(year)
Study
design N (condition) Type participant
%
Male
Mean Age
(SD) Gratitude measure(s) Dependent variable(s) Intervention Time frame ES Summary of ndings
Gordon
(2012)
L 78 (T) College students 17 21 (2.51) RG (AIR) Daily measures of
appreciation,
responsiveness, and
relationship
satisfaction
. . . 2 weeks Not able to calculate Individuals who felt more
appreciated by their partners
reported being more
appreciative of them, and
these appreciative feelings
were associated with greater
responsiveness to a partners
needs.
Gordon
(2012)
L 99 (T) College students 16 20 (2.00) RG (AIR) Relationship commitment;
Daily measures of
appreciation and
relationship
commitment
. . . 1 week and follow-up
at 9 months
Not able to calculate Individuals became more
appreciative of their partners
when they felt appreciated by
them. In turn, individuals who
were more appreciative of
their partners were more likely
to take the risky step of
maintaining their commitment
to their relationships over
time: Appreciation inuences
not just how people think and
act in their relationship, but
also whether they actually
remain in their relationships
over time.
(Continued)
THE JOURNAL OF POSITIVE PSYCHOLOGY 27
Table 5. (Continued).
First
author
(year)
Study
design N (condition) Type participant
%
Male
Mean Age
(SD) Gratitude measure(s) Dependent variable(s) Intervention Time frame ES Summary of ndings
Gordon
(2012)
E 49 couples (T) Young adults 50 24 (6.70) RG (AIR) Relationship satisfaction 6 conversations where
partners take turn;
Observation
. . . Not able to calculate Appreciation is associated with
observer ratings of
responsiveness and
commitment as partners
interact in the laboratory, all
the ndings remained
signicant after controlling for
relationship satisfaction;
Individuals felt more
appreciated by partners who
were seen by observers as
being committed and
responsive to their partners
needs. These behavioral
displays were one way in
which appreciation was
communicated between
partners. When one partner
felt appreciative and engaged
in maintenance behaviors that
transmitted his or her
appreciation, the other
partner felt more appreciated.
Feelings of appreciation seem
to create an upward cycle
whereby appreciation
promotes relationship
maintenance and relationship
maintenance promotes
appreciation.
(Continued)
28 L. JANS-BEKEN ET AL.
Table 5. (Continued).
First
author
(year)
Study
design N (condition) Type participant
%
Male
Mean Age
(SD) Gratitude measure(s) Dependent variable(s) Intervention Time frame ES Summary of ndings
Joel
(2013)
E 216 (T)
NR (EG)
NR (ACG)
NR (CG)
Adults in
a romantic
relationship
40 30 (12.00) RG (3 items for
feeling grateful for
partner)
Relationship commitment Recalling partners
investment, own
investment or no recall
. . . Not able to calculate Individuals who recalled their
romantic partners past
investments felt more
committed to their
relationship, relative to those
who recalled their own
investments or to those in the
no-recall control condition;
Two mediators were found of
this association. When
individuals thought about
their romantic partners
investments into the
relationship, they experienced
greater feelings of trust, which
in turn
predicted stronger feelings of
commitment. Recalling the
romantic partners
investments elicited feelings
of gratitude, which in turn
increased individuals own
commitment to the
relationship.
Joel
(2013)
L 36(T) College students 16 20 (2.00) RG (1 item for feeling
grateful for
partner)
Relationship commitment Daily diary (1 week) 9 months Not able to calculate Perceiving a partner as being
highly investing in
a relationship promotes
commitment. Individuals who
perceived their partners more
investing across 7 days
experienced increased
commitment to their
relationship 9 months later;
This increase in commitment
was due to individuals feeling
more grateful for partners
whom they perceived as more
investing.
(Continued)
THE JOURNAL OF POSITIVE PSYCHOLOGY 29
Table 5. (Continued).
First
author
(year)
Study
design N (condition) Type participant
%
Male
Mean Age
(SD) Gratitude measure(s) Dependent variable(s) Intervention Time frame ES Summary of ndings
Joel
(2013)
L 69 couples (T) Adults 50 NR RG (AIR) Relationship commitment Daily diary (2 weeks) 3 months Not able to calculate The more frequently individuals
thought that their partner
invested into the relationship
over the course of 2 weeks,
the greater the increase in
commitment over a 3-month
period of time. Own
investment frequency
did not signicantly predict
commitment to the
relationship 3 months later;
Individuals who thought that
their partners invested a great
deal into their relationships
over 2 weeks felt more
grateful toward their partner 3
months later, which in turn
lead
to increased commitment to the
relationship.
Kubacka
(2011)
L 195 couples (T) Newly weds 50 Men
32 (4.86)
Women
29 (4.28)
RG (adapted GQ6) Partner responsiveness;
Relationship
maintenance;
Relationship
satisfaction
. . . 2 years and 9 months
after the wedding
Not able to calculate Gratitude is a signal for perceived
partner responsiveness and
a motivator for relationship
maintenance behaviors; There
are both interpersonal and
intrapersonal eects; The
dyadic model held at three
time points separated by
intervals of about 1 year. After
4 years into marriage not only
does the experience of
gratitude motivate the self to
maintain the relationship but
also these relationship
maintenance behaviors are
noticed by the partner who
perceives the self to be
responsive to his or her needs,
and in turn experiences
gratitude.
Lambert
(2011)
L 179 (T) College students 21 NR GB (3-item measure
of behavior)
Comfort of voicing
relationship concerns
. . . 3.5 months β= .18 (p< .01) Expressing gratitude predicts
comfort with relationship
concerns on the long run.
(Continued)
30 L. JANS-BEKEN ET AL.
Table 5. (Continued).
First
author
(year)
Study
design N (condition) Type participant
%
Male
Mean Age
(SD) Gratitude measure(s) Dependent variable(s) Intervention Time frame ES Summary of ndings
Lambert
(2011)
E 71 (EG)
78 (ACG)
76 (CG)
College students 11 20 (9.50) . . . Comfort of voicing
relationship concerns
Gratitude letter (once) . . . d= .39 (p< .05) Experimentally manipulated
expression of gratitude
increased participantscomfort
in voicing relationship
concerns.
Lambert
(2011)
E 18 (EG)
17 (ACG)
20 (ACG)
19 (ACG)
College students 20 19 (1.25) . . . Comfort of voicing
relationship concerns
Expressing gratitude 3 weeks d= .62 (p < .05) Expressing gratitude predicts
comfort with relationship
concerns on the long run and
this association is mediated by
the positive perception of the
partner.
Ng (2017) E 107 (EG)
105(ACG)
College students 33 21 (1.86) SG (GAC) Conformity Writing about gratitude
before rating colors on
a sheet with correct or
bogus answers from
fake participants.
... d= .33 (p< .05) The participants in the gratitude
condition were more likely to
show conformity than the
participants in the neutral
condition, adjusted for
positive aect.
Ng (2017) E 111 (gratitude)
110 (joy)
110 (neutral)
Adults 76 31 (8.40) SG (GAC) Conformity Writing about gratitude
before choosing
between 2 products
incl. their market
shares.
... η
2
= .02 (p< .05) The participants in the gratitude
condition were more likely to
show conformity than the
participants in the joy and
neutral condition, adjusted for
positive aect.
Williams
(2015)
E 30 (EG)
40 (CG)
College students 23 19 (1.50) . . . Aliation with unknown
peer
Receiving feedback with
or without gratitude
1 week Not able to calculate The observed increase in
aliation directed toward
grateful individuals is
mediated by higher
perceptions of interpersonal
warmth resulting from the
expression of gratitude.
Note. L = longitudinal; E = experimental; T = total group, EG = experimental group; CG = control group; ACG = active control group; NR = not reported; RG = relational gratitude; SG = state; GB = gratitude behaviour;
GQ6 = Gratitude Questionnaire 6; GAC = Gratitude Adjectives Checklist; AIR = Appreciation In Relationships, ES = eect size, * = estimation based on results in article.
THE JOURNAL OF POSITIVE PSYCHOLOGY 31
participation on the topics of (i) relationships and (ii)
prosocial behaviour. Table 5 presents an overview of
included studies.
Relationships
Prospective observational as well as RCT work by
Lambert and Fincham (2011) has shown gratitude to
predict comfort with voicing future relationship con-
cerns in close relationships, an association mediated
by a positive perception of the partner (Lambert &
Fincham, 2011). Feeling appreciated, furthermore, eli-
cits appreciating behaviour, relationship maintenance
behaviour, and responsiveness to the partner as
demonstrated both by prospective observational as
well as experimental studies (Algoe & Zhaoyang,
2016;Gordon,Impett,Kogan,Oveis,&Keltner,2012;
Kubacka, Finkenauer, Rusbult, & Keijsers, 2011).
Observational studies by Joel, Gordon, Impett,
MacDonald, and Keltner (2013) have shown, further-
more, that the perception of the partners investment
in the relationship increases feelings of gratitude
which in turn increase relationship commitment
over time. The randomized experimental studies by
Cho and Fast (2012) suggest that, in relationships
involving a hierarchical imbalance, such as in work-
place contexts, expressing gratitude by a subordinate
can ameliorate the tendency to denigrate the com-
petency of a subordinate by a supervisor, because of
an increased sense of social worth for the subordi-
nate and a decreased perceived threat to the own
competency. Showing gratitude may improve subor-
dinate-supervisor relationships, although ongoing
gratitude expression from a subordinate to
a supervisor may signal inferiority, thereby maintain-
ing any existing hierarchical imbalance (Cho & Fast,
2012). Caleon et al. (2017) investigated relatedness of
students towards parents, teachers, and peers. After
a socially oriented gratitude intervention, students in
the experimental (vs. waitlist control) group reported
increased relatedness towards parents and peers who
were more close to them, but not towards teachers.
Regarding the formation of new relationships, experi-
mental work by Williams and Bartlett (2015)suggests
that expressing gratitude facilitates aliation
between unknown peers, and the perception of inter-
personal warmth of the expresser plays a pivotal role
in forming new relationships. Lastly, Diebel,
Woodcock, Cooper, and Brignell (2016), in a school-
based gratitude diary intervention study with random
group assignment, found primary school children in
the gratitude intervention vs. neutral events group to
show an improved sense of belonging (psychological
membership), with boys beneting more from the
intervention than girls.
Prosocial behaviour
A series of experiments performed by Grant and Gino
(2010), in which a (manipulated) written expression of
gratitude motivated beneciaries to assist both the
benefactor as well as a third person, have demonstrated
that gratitude can spark upstream reciprocity, i.e.
returning kindness not only to the benefactors but
also to other parties (Nowak & Roch, 2007). Upstream
reciprocity was also observed in the prospective obser-
vational study by Froh, Bono, and Emmons (2010): gra-
titude predicted social integration, an eect that was
mediated by prosocial behaviour and life satisfaction.
Moreover, gratitude and social integration were found
to serially enhance each other in an upward spiral. Ng
et al. (2017) have linked gratitude to social conformity
based on their nding that experimentally induced gra-
titude in college students and adults raised the like-
lihood of showing private conformity in a colour
judgment task and a material consumption task.
Highly grateful individuals showed more social confor-
mity even when they were making their choices pri-
vately; they chose the wrong answer because they
knew others chose this answer before them (J.W. Ng
et al., 2017). Converse and Fishbach (2012) experimen-
tally dissected the time course of gratitude in response
to prosocial behaviour. Whereas individuals who
receive help from a benefactor in completing a task,
appreciate the assistance more and feel more indebted
during the task than after the task is completed or after
the benefactor is deemed no longer instrumental, ben-
efactors expect to be more appreciated after the task is
nished. Thus, helpers are more appreciated while they
are useful, but do not intuit this eect of task comple-
tion (Converse & Fishbach, 2012).
In conclusion
The majority of the reviewed studies, both prospective
and experimental, suggest that gratitude plays a role in
maintaining healthy relationships, as well as in facilitat-
ing the formation of new relationships. Experimental
and observational work suggests that gratitude
increases prosocial behaviour, not just towards the ben-
efactor but also towards others. This may set in motion
an upward spiral towards positive social behaviour,
reected by improved relationship related emotions,
thoughts, and behaviours benecial for all partners
involved. However, the ndings from Cho and Fast
(2012) suggest that within relationships with an hier-
archical imbalance, gratitude, especially when
expressed naively or excessively, may detrimentally
32 L. JANS-BEKEN ET AL.
impact social relationships by stimulating, rather than
discouraging, feelings of superiority or subordination.
General discussion
With this updated review, we aimed to summarize the
current research regarding state and trait gratitude
associated with human health. Insight in the associa-
tions between gratitude and both physical and mental
health is essential to understand the role of gratitude in
human health and to develop and employ interven-
tions that target those domains in which gratitude can
be expected to contribute to the enhancement of an
individuals health.
Based on our review of the literature, we concluded
that (i) there is currently little convincing evidence for
unique benecial eects of gratitude on physical health
and bodily functions; (ii) having a grateful disposition is
positively linked to the absence of psychopathology,
but gratitude interventions are not unequivocally estab-
lished as universally eective for decreasing psycho-
pathological symptoms; (iii) although a sense of
gratitude seems closely tied to the concept of psycho-
logical well-being, the literature on the impact of grati-
tude (interventions) on psychological well-being
remains scant, fragmented, and inconclusive; (iv) grati-
tude is positively associated with emotional well-being,
and gratitude interventions hold potential for moder-
ately increasing aspects thereof; and (v) gratitude
appears to facilitate social well-being.
The pattern of observations suggests gratitude (inter-
ventions) to moderately benet factors of mental well-
being (emotional well-being, social well-being, and to
a lesser extent psychological well-being), but not
necessarily reduce symptoms of psychopathology.
These ndings align with recent prospective observa-
tional work from Jans-Beken et al. (2017), demonstrating
trait gratitude to predict the presence of future subjec-
tive well-being but not the absence of psychopathology.
Moreover, a series of meta-analyses recently conducted
by Dickens (2017), suggest that gratitude interventions
can benet individual subjective well-being, happiness,
life satisfaction, and positive aect, but their eects on
depression, stress and negative aect are equivocal.
According to Keyestwo-continua model (Keyes, 2002,
2005), mental well-being and psychopathology are two
related but distinct dimensions of complete mental
health (Lamers, Westerhof, Glas, & Bohlmeijer, 2015), i.e.
the presence of mental well-being does not necessarily
imply the absence of psychopathology and vice versa.
Findings from our review thus suggest gratitude (inter-
ventions) to most likely aect the mental well-being
rather than psychopathology dimension of mental
health. However, small to moderate mediating associa-
tions were established for gratitude and meaning in life
on depressive symptoms (Disabato et al., 2017), the
synergy of gratitude and grit, and meaning in life on
suicide ideation (Kleiman et al., 2013), gratitude and
empathy on aggression (DeWall et al., 2012), and grati-
tude and deliberate rumination on post-traumatic
growth (Zhou & Wu, 2015). Given the interrelatedness
of both continua of mental health (Lamers et al., 2015),
cultivating a sense of gratitude may thus indirectly
decrease psychopathology through increasing levels of
mental well-being. In any case, gratitude shows complex
connections with the presence of mental well-being and
absence of psychopathology, that should be taken into
consideration when studying the dynamics of gratitude
and mental health (Jans-Beken et al., 2017), which is
important because mental health and physical health
are reciprocally interconnected ((Ohrnberger, Fichera, &
Sutton, 2017)).
Mental well-being in the two-continua model is com-
posed of three factors: emotional well-being, psycholo-
gical well-being, and social well-being. The current
ndings identied the emotional well-being and social
well-being as most susceptible to the potential bene-
cial eects of state and trait gratitude and gratitude-
based interventions, in line with the broaden-and-build
(Fredrickson, 2001) and the nd-remind-and-bind the-
ory (Algoe, 2012). A considerable amount of evidence
suggests interventions such as gratitude journaling,
carried out over a considerable period of time, to ben-
ecially aect emotional well-being parameters,
although with small to moderate eects, and uncer-
tainty about long-term sustainability. The nding that
gratitude appears to play an important role in forming
and maintaining healthy relationships contrasts with
the observation that virtually all gratitude intervention
protocols are directed at the individual rather than
interpersonal level of experience. Protocols to promote
the eects of gratitude within relationships are scarce,
but show promising results (Algoe & Zhaoyang, 2016;
Joel et al., 2013; Kubacka et al., 2011), and could repre-
sent a starting point for developing and testing stan-
dardized intervention protocols for couples, teams,
institutions, and even larger communities, ideally set-
ting in motion an upward spiral of positive social beha-
viour. However, future research should not only aim at
the benets of gratitude within social interactions, as
pointed out by Lavelock et al. (2016), but should also
direct attention to the hindrance or harm gratitude may
have or cause in relationships. Manipulation or exploi-
tation may occur in relationships between individuals
high in trait gratitude. Because of intense feelings of
gratitude for benets received in the past, they may feel
THE JOURNAL OF POSITIVE PSYCHOLOGY 33
obliged to stay in a relationship or may have diculties
establishing boundaries, with possible negative eects
on well-being. Other known key relationship variables
such as assertiveness or a lack thereof (Van Tongeren,
Davis, & Hook, 2014)may relate to the expression of
gratitude and therefore deserve attention in future
research. In addition, future studies should further elu-
cidate the eects of gender, kinship distance, and rela-
tional familiarity on gratitude experience and
expression in social relationships.
Cultivating the grateful trait may help to build
resilience for mental health problems through an
increased mental well-being. Partial support for this
notion comes from a number of studies reporting
improvements in mental well-being in response to
gratitude interventions (Chan, 2011;Chengetal.,
2015;Jackowskaetal.,2016;Krentzmanetal.,
2015; Otto et al., 2016;Ramírezetal.,2014;
Toepfer et al., 2012; Watkins et al., 2015;Wong
et al., 2016). However, an almost equal amount of
studies reported no eects of gratitude interventions
on stress, depression and anxiety (Baxter et al., 2012;
Chan, 2011; Flinchbaugh et al., 2012;Kerretal.,
2015; Khanna & Singh, 2016;Martínez-Martíetal.,
2010;OLeary & Dockray, 2015; Otto et al., 2016;
Owens & Patterson, 2013). Although methodological
inconsistencies addressed in the next paragraph
may partly underlie the mixed results between gra-
titude and mental well-being, they cannot fully
explain the heterogeneity in ndings. Indeed,
research suggests that positive psychology interven-
tions are not always suitable, in particular for indivi-
duals with mental health issues, and the
eectiveness of an intervention is dependable on
psycho-contextual factors such as stress and adver-
sity (Lies et al., 2014;Parks&Biswas-Diener,2013),
as well as patient characteristics (Sergeant &
Mongrain, 2011;Sin,DellaPorta,&Lyubomirsky,
2011). Nonetheless, cultivating a sense of gratitude
has been suggested to aid in preventing mental
problems following adversity (Lies et al., 2014;
Parks & Biswas-Diener, 2013).
The current literature review provides limited con-
vincing evidence for benecial eects of gratitude on
physical health, in line with ndings from a recent
meta-analysis, reporting no substantive eects of grati-
tude interventions on physical health, sleep, and exer-
cise (Dickens, 2017). Given the observation that
gratitude interventions positively aect subjective well-
being, and the well-substantiated notion that happy
people live longer(i.e. high mental well-being is linked
to better health and longevity; Diener & Chan, 2011),
there is a possibility that gratitude interventions may
indirectly and positively impact physical health through
their eects on mental well-being.
Methodological limitations of included studies
Our review of the literature identied a number of con-
cerns regarding methodological aspects of the studies
that were analysed. First, a substantial amount of studies
employed small samples, making them susceptible to
Type-II error (Rosner, 2010). With respect to the ndings
on gratitude and mental well-being, smaller scale studies
have more often yielded negative results than experi-
ments performed on a larger scale, suggesti