Getting to the Heart of the Matter in Later Life: The Central Role of Affect in Health Message Framing

Abstract

Background and objectives:
Adopting healthy behaviors is often influenced by message framing; gain-framed messages emphasize the benefits of engaging in a behavior, whereas loss-framed messages highlight the consequences of not engaging in a behavior. Research has begun to uncover the underlying affective pathways involved in message framing. In the current study, we examined the role of affect in message framing to encourage physical exercise program enrollment among older adults.

Research design and methods:
We mailed flyers to 126 volunteers assigned to a gain- or loss-framed condition and measured their affective reactions to the flyer and enrollment intentions. After the call, participants had the opportunity to contact us to enroll.

Results:
Gain- versus loss-framing led to more positive affect toward the flyer, which predicted intentions and enrollment. In indirect-effect analyses, frame indirectly influenced intentions and enrollment via positive affect.

Discussion and implications:
Although message framing plays an indirect role in influencing behavior, affect plays a central role.

Full text

1
The Gerontologist
cite as: Gerontologist, 2020, Vol. XX, No. XX, 1–7
doi:10.1093/geront/gnaa128
Advance Access publication September 11, 2020
© The Author(s) 2020. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved.
For permissions, please e-mail: journals.permissions@oup.com.
Research Article
Getting to the Heart of the Matter in Later Life: The Central
Role of Affect in Health Message Framing
Joseph A. Mikels, PhD,1,*, Nathaniel A. Young, MA,1 Xiaomei Liu, MS,2, and
ElizabethA.L. Stine-Morrow, PhD2
1Department of Psychology, DePaul University, Chicago, Illinois. 2Department of Educational Psychology and Beckman
Institute, University of Illinois at Urbana-Champaign.
*Address correspondence to: Joseph A.Mikels, PhD, Department of Psychology, DePaul University, 2219N. Kenmore Ave., Chicago, IL 60614.
E-mail: jmikels@depaul.edu
Received: June 17, 2020; Editorial Decision Date: August 26, 2020
Decision Editor: Suzanne Meeks, PhD, FGSA
Abstract
Background and Objectives: Adopting healthy behaviors is often inuenced by message framing; gain-framed messages
emphasize the benets of engaging in a behavior, whereas loss-framed messages highlight the consequences of not engaging
in a behavior. Research has begun to uncover the underlying affective pathways involved in message framing. In the current
study, we examined the role of affect in message framing to encourage exercise program enrollment among older adults.
Research Design and Methods: We mailed yers to 126 volunteers assigned to a gain- or loss-framed condition and
measured their affective reactions to the yer and enrollment intentions. After the call, participants had the opportunity to
contact us to enroll.
Results: Gain versus loss framing led to more positive affect toward the yer, which predicted intentions and enrollmenteffort.
In indirect effect analyses, frame indirectly inuenced intentions and enrollmenteffort via positive affect.
Discussion and Implications: Although message framing plays an indirect role in inuencing behavior, affect plays a central
role.
Keywords: Aging, Decision making, Emotions, Health, Message framing
Theoretical perspectives emphasizing the role of emotions
in decision making are starting to challenge long-
held assumptions from a historical focus on cognitive
explanations; decision making depends not only on cog-
nitive processes and evaluations but, perhaps more signi-
cantly, on emotional reactions as well (Lerner etal., 2015;
Loewenstein etal., 2001; Slovic etal., 2005). The role of af-
fect as an integral source of information in decision making
can be aptly considered within the larger context of dual-
process theories that draw the distinction between two ge-
neral processing streams: intuitive and deliberative (Epstein,
1994; Kahneman, 2003). One dual-process perspective,
cognitive-experiential self-theory (CEST), includes affect as
a central component of the intuitive system (Epstein, 1994).
Specically, according to CEST, behavior and decisions are
generally guided by integrating affect-laden experiential
and rational-analytic streams of information.
Importantly, affective relative to cognitive explanations
also appear to have surprising utility in understanding
adult life-span differences in decision making generally
(Mikels et al., 2015; Peters etal., 2007). Understanding
decision-making processes in later life has major
implications given that the proportion of older adults in
Western societies is approaching unprecedented levels. For
instance, in the United States, nearly 20% of the popula-
tion will be aged 65 or older by 2030 (United States Census
Copyedited by: oup
Downloaded from https://academic.oup.com/gerontologist/advance-article/doi/10.1093/geront/gnaa128/5904202 by DePaul University user on 28 January 2021

2 The Gerontologist, 2020, Vol. XX, No. XX
Bureau, 2012). Largely because of this demographic shift,
health expenditures are approaching 20% of the U.S.gross
domestic product (Cuckler etal., 2013). As greater num-
bers of older adults are faced with making important health
decisions, understanding the most effective strategies for
increasing healthy behaviors—and the underlying causal
pathways—has never been more critical. The current study
was designed to examine the affective mechanisms under-
lying health behavior decisions.
Given the impactful role of emotion in decision making,
an understanding of this relationship can be used to more ef-
fectively tailor health messages to facilitate behavior change.
Message framing has been examined in numerous health
domains including disease prevention (e.g., physical ac-
tivity and diet) and disease detection (e.g., breast cancer and
skin cancer; Rothman & Salovey, 1997). In particular, gain-
framed messages that emphasize the benets of engaging in
a specic behavior are theorized to be more inuential for
prevention relative to loss-framed messages that highlight the
consequences of not engaging in a particular behavior, which
are thought to be most inuential for detection (Rothman
& Salovey, 1997). According to the framework of Rothman
and Salovey (1997), gain-framed messages can either take
the form of acquiring desirable outcomes or avoiding un-
desirable outcomes and vice versa for loss-framed messages.
Recent meta-analyses have provided partial support for the
greater impact of gain-framed versus loss-framed messages in
encouraging preventative behaviors—such as exercise—but
the effects are inconsistent (Gallagher & Updegraff, 2012;
O’Keefe & Jensen, 2006, 2007). The tenuous link between
message framing and behavior reects the multifaceted and
complex causal pathways from framing to behavior including
mediators such as attitudes about the behavior, perceived ef-
fectiveness of the messages, and intentions to engage in the
behavior (Dillard etal., 2007; van’t Riet etal., 2010).
Although these cognitive mediators have underscored
the complexity of message-framing effects, only recently
have affective pathways been considered. In particular, it
is becoming increasingly clear that gain- versus loss-framed
messages evoke different affective reactions, such that gain-
framed messages evoke positive affect, whereas loss-framed
messages evoke negative affect (Liu et al., 2019; Mikels
etal., 2016; van’t Riet etal., 2010). Importantly, the greater
positive affect evoked by gain-framed messages is associ-
ated with higher perceived effectiveness of the messages,
especially for older adults (Liu etal., 2019). Moreover, ev-
idence suggests that gain-framed messages lead to higher
positive affect, greater information acceptance, and more
positive attitudes (van’t Riet et al., 2010). But do these
changes in perspective lead to changes in actual behavior?
The current study sought to address this question through
a eld study using message framing to increase actions to-
ward healthy behavior in a naturalistic setting.
One preventative health domain in which decisions
can have impactful consequences is exercise, especially for
older adults. It is well known that regular exercise improves
physical health for people of all ages (Emery & Gatz, 1990;
Hillman et al., 2008). However, this is especially true for
older individuals who generally experience disproportionate
physical and cognitive declines relative to younger adults
(Prakash etal., 2015); not only is exercise linked to improved
cardiovascular function, greater muscle strength, and better
balance for older adults, but there are also improvements
in executive control processes (Colcombe & Kramer, 2003;
Cress etal., 1999; Howe etal., 2011; Hughes etal., 2004;
Liu & Latham, 2009; Sherrington etal., 2011; Whitehead &
Blaxton, 2017). Despite strong evidence for the importance
of regular physical activity, 40.3% of younger adults (aged
18–44) and 59.4% of older adults (aged older than 65)fail to
meet the recommended levels of aerobic exercise and muscle
strengthening (National Center for Health Statistics, 2015).
Thus, one of the most important preventive health domains
for message framing to target is physical activity. Notably,
gain-framed messages have been shown to increase walking
for older adults (Notthoff & Carstensen, 2014). However,
the mechanisms resulting in this behavioral effect are not
entirely clear. Extant evidence suggests that older adults
nd gain- versus loss-framed messages more motivating
(Notthoff etal., 2016) and that older adults better remember
gain- versus loss-framed messages (Notthoff et al., 2016;
Shamaskin etal., 2010). Thus, motivation and memory likely
play an underlying role in the behavioral effects of message
framing. Nonetheless, given that gain-framed messages elicit
positive affect, which is associated with higher perceived ef-
fectiveness (Liu et al., 2019), important yet unknown key
mechanisms operating in the effects of message framing on
behavior may be affective innature.
The current study was designed to examine how affec-
tive reactions to preventative health messages inuence
intentions and actual behavior. We focused on exercise,
given its broad and profound inuence on physical and
psychological health. Moreover, we focused on older adults
insofar as they are most to benet from exercise and show
a preference toward positive messages (for a review, see
Mikels etal., 2015). We predicted that gain-framed health
messages would have their effect on intentions and actual
behavior via positive affective responses, beyond com-
monly measured cognitive evaluations such as perceived
effectiveness (Dillard etal., 2007).
Method
Participants
The sample consisted of 126 participants (M age=71.75,
SD = 7.83, 69.8% women), who were recruited gen-
erally to participate in our research studies from the
Chicago area via a newspaper advertisement in two
local newspapers with a total circulation of approxi-
mately 200,000 readers. If enrolled in the current study,
they were compensated with a $30 gift card. The experi-
ment was approved by DePaul University’s Institutional
Review Board. Table 1 presents data regarding the
Copyedited by: oup
Downloaded from https://academic.oup.com/gerontologist/advance-article/doi/10.1093/geront/gnaa128/5904202 by DePaul University user on 28 January 2021

The Gerontologist, 2020, Vol. XX, No. XX 3
demographic, health, and affective characteristics of the
participants in each condition. The two groups differed
in positive and negative state affect, p <.05.
Materials
Exercise programyers
The study used two flyers advertising an exercise pro-
gram for older adults: a gain-framed flyer and a loss-
framed flyer (Figure 1). The gain-framed flyer used
benefit-focused messaging, which emphasized the
benefits of engaging in exercise. For instance, the focal
message, “Exercising will improve your health,” was
followed by several additional gain-framed items. The
loss-framed flyer used consequence-focused messaging,
which highlighted the consequences of not exercising.
For instance, the focal message, “Not exercising will
increase health problems,” was followed by the conse-
quential alternatives to the original messages.
Current stateaffect
To measure state affect, we adapted the modied differen-
tial emotional scale (mDES; Fredrickson etal., 2003). The
scale is a measure of current state affect with 11 positive
emotions (amusement, awe, compassion, contentment,
gratitude, hope, interest, joy, love, pride, and surprise)
and 8 negative emotions (anger, contempt, disgust, em-
barrassment, fear, guilt, sadness, and shame). Participants
respond as to the extent they are experiencing each emo-
tion/feeling at the present moment on a scale of 1 (not at
all) to 5 (extremely). Aparticipant’s positive and negative
emotional state was calculated by averaging the 11 pos-
itive emotions and the 8 negative emotions, respectively.
Immediate affective reactions toward theyer
Participants were asked to rate how they felt about three
separate elements of the yer (as demarcated in Figure1):
(a) the text in the red box, (b) the three sentences in the
second section, and (c) the statements in the third section.
Each element was rated on a 1 (very negative) to 6 (very
positive) scale. Overall feelings toward the yer were cal-
culated by averaging the ratings across the three measures.
Flyer perceived effectiveness
Participants were asked to rate how effective the three yer
elements were in convincing them to exercise (as described
above and demarcated in Figure1). Each participant rated
how effective the yers were on a 1 (very ineffective) to 6
(very effective) scale.
Enrollment intentions
Participants were asked to rate how interested they were in
joining the exercise program after the fourth section of the
yers (see Figure 1). Each participant rated how interested
they were on a 1 (not at all) to 5 (extremely) scale.
Behavioral action toward enrollment (enrollmenteffort)
To measure behavioral action, participants were given an
opportunity to take action to enroll in the exercise pro-
gram. They were able to either mail in a postcard or enter
their name in a secure website. Action to enroll was either
coded “1” for participants who took action and “0” for
participants who did not.
Procedure
Upon expressing interest in the study, potential participants
answered calls by the researchers, and the study was
conducted in two parts by phone. First, the participants
completed an initial phone screening and we obtained
verbal informed consent. If they agreed to participate,
participants set up a schedule to receive a mailed packet
(containing the yer as well as a postcard) and a follow-up
phone call approximately a week following the mailing
of the packet. We emphasized in the instructions during
the call that they were not to open the envelope until they
Table 1. Participant Characteristics by Frame Condition
Gain, N=63 Loss, N=63 Statistic
M (SD)M (SD)t p
Age (in years) 71.92 (7.90) 71.56 (7.76) 0.26 .79
Sex 37.3% F, 12.7% M 32.5% F, 17.5% M 1.16 .25
Highest level of education 7.95 (1.88) 7.91 (1.89) 0.14 .89
Scaled income 10.81 (8.56) 9.97 (8.63) 0.55 .58
Physical health 47.73 (11.04) 48.98 (9.14) 0.69 .49
Mental health 50.58 (11.78) 53.05 (7.98) 1.38 .17
Positive state affect 2.95 (0.80) 3.22 (0.70) −2.05 .042
Negative state affect 1.64 (0.65) 1.41 (0.51) 2.23 .028
Notes: Sex: F=female, M=male; highest level of education: on a scale of 1–11 (with 1 indicating “No schooling completed” and 11 indicating “Doctorate de-
gree”); income: on a scale of 1–12 (with 1 indicating “Less than $10,000 and 12 indicating “More than $150,000”); physical and mental health norm-based com-
posite scores from the SF-36; positive state affect: an average of 10 positive emotions on a scale of 1–5 (with 1 indicating “not at all” and 5 indicating “extremely”);
negative state affect: an average of 9 negative emotions on a scale of 1–5 (with 1 indicating “not at all” and 5 indicating “extremely”).
Copyedited by: oup
Downloaded from https://academic.oup.com/gerontologist/advance-article/doi/10.1093/geront/gnaa128/5904202 by DePaul University user on 28 January 2021

4 The Gerontologist, 2020, Vol. XX, No. XX
received the second call. Additionally, we sealed the enve-
lope with a large label stating in large bold font to not open
the envelope until they received a call from us. During the
second phone call, participants rst completed the mDES,
opened the packet, then read the sections of the yer to
the experimenter, were asked how they felt about and their
perceived effectiveness ofthe three sections of the yer, and
nally after reading section four were asked about their
intentions in joining the exercise program. They were then
given an opportunity to send in a postcard or visit a web-
site to take action toward joining the exercise program.
Results
Our analyses focused on the role of immediate affective
reactions relative to cognitive evaluations in the inuence of
gain- versus loss-framed messages on intentions to enroll in
the exercise program as well as taking action to enroll in the
exercise program. First, we examined the inuence of gain-
versus loss-framed messages on immediate affective reactions
toward the yer, perceived effectiveness of the yer, intentions
to enroll in the exercise program, and action to enroll in the
exercise program. Next, analyses examined whether imme-
diate affect toward the yer and/or perceived effectiveness
predicted intentions in the exercise program and enrollment
effort to join the exercise program. Finally, two mediation
analyses examined if there were indirect effects (IEs) of gain-
versus loss-framed messages on intentions in the exercise
program and action to enroll in the exercise program via
immediate affect toward the yer (controlling for perceived
effectiveness). Given the group differences in incidental state
affect, all analyses were conducted with and without control-
ling for positive and negative state affect as measured by the
mDES. State affect did not change the pattern of results re-
garding the effect of frame on immediate affective reactions,
effectiveness, or the IEs of frame on intentions and effort to
enroll. As such, the following analyses are reported without
controlling for incidental stateaffect.
The Influence of Frame on Immediate Affect,
Effectiveness, Intentions, and Action toEnroll
A multivariate analysis of variance was conducted to
examine the effect of frame on immediate affect to-
ward the yer, perceived effectiveness of the yer, and
intentions to enroll in the exercise program; for means,
see Table 2. The analysis revealed a signicant effect of
gain- versus loss frame on immediate affect toward the
yer (F(1, 124)= 16.36, p < .001, ɳ2= 0.117), such that
gain frames evoked more positive affect than loss frames.
Considering possible sex differences in affective reactions,
immediateImmediate affective reactions did not differ by
Gain-framed Flyer Loss-framed Flyer
Section
One
Section
Two
Section
Three
Section
Four
Figure 1. Exercise program flyers.
Copyedited by: oup
Downloaded from https://academic.oup.com/gerontologist/advance-article/doi/10.1093/geront/gnaa128/5904202 by DePaul University user on 28 January 2021

The Gerontologist, 2020, Vol. XX, No. XX 5
sex. Regarding the other measures, frame did not have an
effect on effectiveness or intentions. Furthermore, a logistic
regression showed that there was no direct effect (DE) of
frame on action toenroll.
Did Immediate Affect or Perceived Effectiveness
Predict Intentions and Action toEnroll?
Four separate regression analyses examined if immediate
affect and/or perceived effectiveness toward the yer
predicted intentions and enrollment effort (note that im-
mediate affect and perceived effectiveness were correlated,
r = 0.60, p < .001). The rst regression examined if im-
mediate affect predicted intentions to enroll in the exercise
program. This analysis revealed that the more positive im-
mediate affect was toward the yer, the greater intentions
there were in the exercise program (β=0.569, SE=0.088, p
< .001, 95% condence interval [CI]: 0.40–0.74, ɳ2=0.25).
Logistic regression was used to examine whether immediate
affect predicted enrollment effort. This analysis revealed
that more positive immediate affect toward the yer signif-
icantly predicted an increased probability of taking action
to enroll in the exercise program (β=0.107, SE=0.044,
p < .02, odds ratio [OR]=1.11, 95% CI: 1.02–1.21). The
third analysis examined whether perceived effectiveness
predicted intentions to enroll in the exercise program. This
analysis revealed that higher perceived effectiveness signif-
icantly predicted increased intentions to enroll in the ex-
ercise program (β=0.629, SE=0.07, p < .001, 95% CI:
0.48–0.77, ɳ2=0.37). Then a logistic regression analysis
examined if perceived effectiveness predicted enrollment
effort. This analysis revealed that perceived effectiveness
marginally predicted increased probability to take action
to enroll in the exercise program (β = 0.355, SE=0.18,
p=.051, OR=1.43, 95% CI: 1.02–2.09).
Did Frame Have an IE on Intentions and
Action to Enroll in the Exercise Program via
ImmediateAffect?
Given that frame had a DE on immediate affect but not
perceived effectiveness, the analyses only examined imme-
diate affect. Two separate IE analyses examined whether
there was an IE of gain- versus loss-framed messages on
intentions to enroll in the exercise program and action
to enroll in the program (Figure 2). Both analyses used
Hayes’ (2009) method for examining IEs. Additionally,
both analyses used the mediation package (Tingley etal.,
2014)in R (R Core Team, 2013)to estimate the causal me-
diation effect (IE) and the DE. Both the IE and DE estimates
Table 2. The Influence of Frame on Affective Reactions, Ratings of Effectiveness and Intentions, and Action toEnroll
Gain, N=63 Loss, N=63 Statistic
M (SD)M (SD) p
Immediate affective reactions 5.41 (0.671) 4.73 (1.16) F16.36 <.001
Perceived effectiveness 5.04 (1.01) 4.95 (1.19) F0.185 .668
Enrollment intentions 3.51 (1.09) 3.38 (1.18) F0.392 .532
Action to enroll 43.0% 49.2% β0.256 .475
Frame Feelings About the
Flyer
Acon to Enroll
Intenons to Enroll
Gain frames led to more positive feelings
toward the flyer than loss frames (= .68,
SE = .17, 95% CI: .35, 1.02 , p< .001)
Greater positive feelings toward the flyer led
to an increased likelihood of taking action to
enroll ( = .11, SE = .04, OR = 1.11,
95% CI:
1.02, 1.21 , p< .02)
Greater positive feelings toward the flyer led
to greater intentions to enroll ( = .57, SE =
.09, 95% CI: .40, .74, p< .001)
Frame indirectlyinfluenced intenons to enroll via the manipulaon of immediate affect experienced
toward the flyer (IE: = .42, 95% CI = .20, .68, p< .001)
Frame indirectlyinfluenced acon to enroll via the manipulaon of immediate affect experienced toward
the flyer (IE: = .09, 95% CI = .02, .16,
p
< .003)
Figure 2. Indirect effect analyses examining the role of immediate affect in the effect of frame on intentions and enrollment effort. Note: These sta-
tistics are reported without controlling for perceived effectiveness.
Copyedited by: oup
Downloaded from https://academic.oup.com/gerontologist/advance-article/doi/10.1093/geront/gnaa128/5904202 by DePaul University user on 28 January 2021

6 The Gerontologist, 2020, Vol. XX, No. XX
were computed using 5,000 bootstrapped samples, and the
95% CI was computed by determining the IE and DE at the
2.5th and 97.5th percentiles.
The rst analysis examined whether frame indirectly
inuenced intentions via immediate feelings toward the
yer. This analysis indicated that there was an IE of frame
on intentions via immediate feelings toward the yer (IE:
β=0.42, 95% CI: 0.20–0.68, p < .001). Compared to the
loss frame, the gain frame was associated with increased
intentions to enroll in the exercise program via more posi-
tive immediate affect toward the yer. The IE remained sig-
nicant even when controlling for perceived effectiveness
(IE: β=0.17, 95% CI: 0.03–0.33, p <.02).
The second analysis examined whether frame indirectly
inuenced action to enroll via immediate feelings toward
the yer. This analysis indicated that there was an IE of
frame on enrollment effort via immediate feelings toward
the yer (IE: β = 0.09, 95% CI: 0.02–0.16, p < .003).
Compared to the loss frame, the gain frame was associ-
ated with an increased likelihood to take action to enroll
in the exercise program via more positive immediate affect
toward the yer. As with the rst analysis, here too, the IE
remained signicant when controlling for perceived effec-
tiveness (β=0.07, 95% CI: 0.001–0.15, p=.041).
Discussion
The ndings of the current eld study revealed a central
role for affect in the inuence of framing on intentions, but
more importantly, on enrollment effort. In particular, mes-
sage framing had IEs on the outcome measures via feelings
about the messages but not via evaluative judgments of
perceived effectiveness. Gain- versus loss-framed messages
evoked more positive feelings about the messages, which in
turn predicted intentions to enroll in the program and be-
havioral action to enroll in the program. These results are
the rst to directly link message framing to actual behavior
via an affective pathway.
This work was conducted in the important preventative
health domain of exercise, for which message framing may
have an impact. Importantly, though, in this study, the actual
frame of the message was not as inuential on intentions
or enrollment effort as the immediate affective reactions
of individuals to the messages. In other words, the greater
the extent to which loss-framed messages elicited positive
affect predicted greater intentions and behavior. As such,
these results suggest that message framing aimed at pre-
ventative behaviors is effective only to the degree to which
it elicits positive affect. It is interesting to note that the im-
mediate affective reactions did not differ by sex, suggesting
that extraneous factors such as sex may not have a strong
effect on affective responses to message framing.
Although these results are compelling, there are caveats.
First, the sample only included older adults, for whom
increasing exercise is especially benecial given their dispro-
portionate physical and cognitive declines relative to younger
adults (Prakash et al., 2015). Given age differences in
motivation and emotion (Carstensen, 2006), it is possible that
individuals of different ages may respond differently. Thus,
examining message framing across the adult life span could
provide additional insights. Also, including other preventative
health domains (e.g., diet and nutrition, safer sex practices,
and dental hygiene) would provide greater generalizability.
Finally, the current nding that frame did not affect perceived
effectiveness deserves further consideration. Whereas Liu
etal. (2019) found that gain-framed messages were perceived
to be more effective than loss-framed messages, this effect
was not found in the current study. However, we did nd that
more positive immediate affect was signicantly associated
with higher perceived effectiveness, partially replicating Liu
etal. (2019). This discrepant pattern may be due to method-
ological differences (e.g., sample and design).
In summary, the current work is the rst to demon-
strate the role of affect in the pathway from message
framing to behavior. Creating preventive messages that
evoke positive affect may be the most effective way for-
ward in efforts to increase healthy behaviors. Ultimately,
the path toward preventive health—as inuenced by
persuasive messaging—appears intrinsically linked with
positiveaffect.
Funding
This research was partially supported by the National Institute
on Aging, grant R01-AG043533 and by the National Science
Foundation, grant SES-1536260.
Conflict of Interest
None declared.
References
Carstensen,L.L. (2006). The inuence of a sense of time on human
development. Science (New York, N.Y.), 312(5782), 1913–1915.
doi:10.1126/science.1127488
Colcombe,S., & Kramer,A.F. (2003). Fitness effects on the cognitive
function of older adults: Ameta-analytic study. Psychological
Science, 14(2), 125–130. doi:10.1111/1467-9280.t01-1-01430
Cress, M. E., Buchner, D. M., Questad, K. A., Esselman, P. C.,
deLateur, B. J., & Schwartz, R. S. (1999). Exercise: Effects on
physical functional performance in independent older adults. The
Journals of Gerontology, Series A: Biological Sciences and Medical
Sciences, 54(5), 242–248. doi:10.1093/gerona/54.5.m242
Cuckler,G.A., Sisko,A.M., Keehan,S.P., Smith,S.D., Madison,A.J.,
Poisal, J. A., Wolfe, C., Lizonitz, J., & Stone, D. A. (2013).
National Health Expenditure Projections, 2012–22: Slow
growth until coverage expands and economy improves. Health
Affairs, 32(10), 1820–1831. doi:10.1377/hlthaff.2013.0721
Dillard,J.P., Weber,K. M., & Vail,R. G. (2007). The relationship
between the perceived and the actual effectiveness of persua-
sive messages: Ameta-analysis with implications for formative
Copyedited by: oup
Downloaded from https://academic.oup.com/gerontologist/advance-article/doi/10.1093/geront/gnaa128/5904202 by DePaul University user on 28 January 2021

The Gerontologist, 2020, Vol. XX, No. XX 7
campaign research. Journal of Communication, 57, 613–631.
doi:10.1111/j.1460-2466.2007.00360.x
Emery,C.F., & Gatz,M. (1990). Psychological and cognitive effects
of an exercise program for community-residing older adults. The
Gerontologist, 30, 184–188. doi:10.1093/geront/30.2.184
Epstein, S. (1994). Integration of the cognitive and the psychody-
namic unconscious. The American Psychologist, 49(8), 709–
724. doi:10.1037//0003-066x.49.8.709
Fredrickson,B. L., Tugade,M.M., Waugh,C.E., & Larkin, G.R.
(2003). What good are positive emotions in crises? A pro-
spective study of resilience and emotions following the ter-
rorist attacks on the United States on September 11th, 2001.
Journal of Personality and Social Psychology, 84(2), 365–376.
doi:10.1037//0022-3514.84.2.365
Gallagher, K. M., & Updegraff, J. A. (2012). Health message
framing effects on attitudes, intentions, and behaviors: A meta-
analytic review. Annals of Behavioral Medicine, 43, 101–116.
doi:10.1007/s12160-011-9308-7
Hayes,A. F. (2009). Beyond Baron and Kenny: Statistical mediation
analysis in the new millennium. Communication Monographs,
76(4), 408–420. doi:10.1080/03637750903310360
Hillman,C.H., Erickson,K.I., & Kramer,A.F. (2008). Be smart, ex-
ercise your heart: Exercise effects on brain and cognition. Nature
Reviews. Neuroscience, 9(1), 58–65. doi:10.1038/nrn2298
Howe,T.E., Rochester, L., Neil,F., Skelton,D.A., & Ballinger,C.
(2011). Exercise for improving balance in older people.
Cochrane Database of Systematic Reviews, 2011(11).
doi:10.1002/14651858.CD004963
Hughes,S.L., Seymour,R.B., Campbell,R., Pollak,N., Huber,G.,
& Sharma,L. (2004). Impact of the t and strong intervention
on older adults with osteoarthritis. The Gerontologist, 44(2),
217–228. doi:10.1093/geront/44.2.217
Kahneman, D. (2003). A perspective on judgment and choice:
Mapping bounded rationality. The American Psychologist,
58(9), 697–720. doi:10.1037/0003-066X.58.9.697
Lerner,J.S., Li,Y., Valdesolo,P., & Kassam,K.S. (2015). Emotion
and decision making. Annual Review of Psychology, 66, 799–
823. doi:10.1146/annurev-psych-010213-115043
Liu, C. J., & Latham, N. K. (2009). Progressive resistance
strength training for improving physical function in older
adults. Cochrane Database of Systematic Reviews, 2009(3).
doi:10.1002/14651858.cd002759.pub2
Liu,X., Shuster,M.M., Mikels,J.A., & Stine-Morrow,E.A.L.
(2019). Doing what makes you happy: Health message framing
for younger and older adults. Experimental Aging Research,
45(4), 293–305. doi:10.1080/0361073X.2019.1627491
Loewenstein,G.F., Weber,E.U., Hsee,C.K., & Welch,N. (2001).
Risk as feelings. Psychological Bulletin, 127(2), 267–286.
doi:10.1037/0033-2909.127.2.267
Mikels, J. A., Shuster, M. M., & Thai, S. T. (2015). Aging, emo-
tion, & decision making. In T. M. Hess, J. Strough, &
C.E.Loeckenhoff (Eds.), Aging and decision making: Empirical
and applied perspectives (pp. 169–188). Elsevier Academic Press.
doi:10.1016/B978-0-12-417148-0.00009-1
Mikels,J.A., Shuster,M.M., Thai,S.T., Smith-Ray,R., Waugh,C.E.,
Roth,K., Keilly, A., & Stine-Morrow, E. A. (2016). Messages
that matter: Age differences in affective responses to framed
health messages. Psychology and Aging, 31(4), 409–414.
doi:10.1037/pag0000040
National Center for Health Statistics. (2015). Health, United States,
2014: With special feature on adults aged 55–64. Author.
Notthoff, N., & Carstensen, L. L. (2014). Positive messaging
promotes walking in older adults. Psychology and Aging, 29(2),
329–341. doi:10.1037/a0036748
Notthoff,N., Klomp,P., Doerwald,F., & Scheibe,S. (2016). Positive
messages enhance older adults’ motivation and recognition
memory for physical activity programmes. European Journal of
Ageing, 13(3), 251–257. doi:10.1007/s10433-016-0368-1
O’Keefe,D.J., & Jensen,J.D. (2006). The advantages of compliance
or the disadvantages of noncompliance? Ameta-analytic review
of the relative persuasive effectiveness of gain-framed and loss-
framed messages. Annals of the International Communication
Association, 30, 1–43. doi:10.1080/23808985.2006.11679054
O’Keefe, D. J., & Jensen, J. D. (2007). The relative persua-
siveness of gain-framed and loss-framed messages for en-
couraging disease prevention behaviors: A meta-analytic
review. Journal of Health Communication, 12(7), 623–644.
doi:10.1080/10810730701615198
Peters,E., Hess,T.M., Västfjäll,D., & Auman,C. (2007). Adult age
differences in dual information processes: Implications for the
role of affective and deliberative processes in older adults’ deci-
sion making. Perspectives on Psychological Science, 2(1), 1–23.
doi:10.1111/j.1745-6916.2007.00025.x
Prakash, R. S., Voss, M. W., Erickson, K. I., & Kramer, A. F.
(2015). Physical activity and cognitive vitality. Annual
Review of Psychology, 66, 769–797. doi:10.1146/
annurev-psych-010814-015249
R Core Team. (2013). R: A language and environment for statistical
computing. https://www.R-project.org/
van’t Riet, J., Ruiter, R. A., Werrij, M. Q., Candel, M. J., &
deVries,H. (2010). Distinct pathways to persuasion: The role
of affect in message-framing effects. European Journal of Social
Psychology, 40, 1261–1276. doi:10.1002/ejsp.722
Rothman,A.J., & Salovey,P. (1997). Shaping perceptions to motivate
healthy behavior: The role of message framing. Psychological
Bulletin, 121(1), 3–19. doi:10.1037/0033-2909.121.1.3
Shamaskin,A.M., Mikels,J.A., & Reed,A.E. (2010). Getting the
message across: Age differences in the positive and negative
framing of health care messages. Psychology and Aging, 25(3),
746–751. doi:10.1037/a0018431
Sherrington,C., Tiedemann,A., Fairhall,N., Close,J.C., & Lord,S.R.
(2011). Exercise to prevent falls in older adults: An updated meta-
analysis and best practice recommendations. New South Wales
Public Health Bulletin, 22(3–4), 78–83. doi:10.1071/NB10056
Slovic,P., Peters,E., Finucane,M.L., & Macgregor,D. G. (2005).
Affect, risk, and decision making. Health Psychology, 24(4S),
S35–S40. doi:10.1037/0278-6133.24.4.S35
Tingley,D., Yamamoto,T., Hirose,K., Keele,L., & Imai,K. (2014).
Mediation: R package for causal mediation analysis. Journal of
Statistical Software. doi:10.18637/jss.v059.i05
U.S. Census Bureau (2012). Percent distribution of the projected
population by selected age groups and sex for the United States:
2015 to 2060 [Data le]. https://www.census.gov/data/ta-
bles/2012/demo/popproj/2012-summary-tables.html
Whitehead,B.R., & Blaxton,J.M. (2017). Daily well-being benets
of physical activity in older adults: Does time or type matter?
The Gerontologist, 57(6), 1062–1071. doi:10.1093/geront/
gnw250
Copyedited by: oup
Downloaded from https://academic.oup.com/gerontologist/advance-article/doi/10.1093/geront/gnaa128/5904202 by DePaul University user on 28 January 2021