Abstract
When people have more negative perceptions about aging or attribute health decline to old age, they engage in less health promotion behavior. We tested whether an intervention of brief anti-ageism messages addressing views of aging could motivate engagement in physical activities at senior centers. Attendees aged 50 and older (n = 349; Mage = 72, SD = 9) at seven centers were randomly assigned to read one of three intervention messages (different approaches addressing views of aging) or to not read a message before rating their likelihood of attending a variety of center programs, including physical activities. Multilevel regression models indicated the intervention increased motivation to attend physical activities compared with the control group among participants aged 72 and older. The three anti-ageism messages were similarly effective suggesting some flexibility in framing. The results indicate anti-ageism messages may be a scalable, low-cost approach to promoting physical activity in older adults.
Age stereotypes and negative perceptions about aging are so common in our society that we can continue to endorse them even after we become older adults ourselves (when they become self-directed stereotypes or self-perceptions of aging; Levy, 2009). Unfortunately, this exposure to and internalization of negative views of aging are linked to a host of negative outcomes as we age, including less engagement in preventive health behaviors (Levy & Myers, 2004; Stewart et al., 2012) and even mortality risk (Sargent-Cox et al., 2014). In fact, enough evidence links views of aging to health and quality of life that intervention is recommended (e.g., Levy, 2017; Nelson, 2016). However, relatively little published research examines how to shift middle-aged and older adults’ views of aging to maintain and promote wellness. In this study, we conducted a field experiment at metropolitan area senior centers testing whether a brief anti-ageism message intervention could increase motivation to attend physical activities and other center programming among middle-age and older adults.
Links Between Views on Aging and Health in Later Life
In this article we use the umbrella term views on aging to encompass attitudes and beliefs about one’s own or others’ aging (e.g., age stereotypes, expectations about aging, and positive and negative perceptions of one’s aging experience—also known as self-perceptions of aging). Over the past 20 years a substantial body of experiments and longitudinal studies indicate that exposure to and endorsement of more negative and less positive views of aging can undermine older adults’ health and quality of life in a variety of domains. Views of aging are linked to physical, cognitive, and even social functioning (e.g., Levy & Leifheit-Limson, 2009; Menkin et al., 2017; Westerhof et al., 2014). Particularly relevant to the current study, less negative and more positive views of aging have been specifically linked to higher levels of physical activity (Sarkisian et al., 2005; Wurm et al., 2010).
Although there are multiple possible pathways for these effects (see Levy, 2009 for review), a central mechanism appears to be that views on aging act as self-fulfilling prophecies. When people have more negative expectations about aging and attribute declines in health and function to age, they are less motivated to engage in behaviors that might ameliorate or help stave off further decline. For example, reducing maladaptive views of aging such as fatalism (i.e., expecting inevitable decline due to aging) is expected to encourage older adults to engage in more physical activity by promoting feelings of control over aging outcomes (Sarkisian et al., 2007) and belief that there is a worthwhile future to prepare for (Wurm et al., 2010).
Interventions to Shift Views of Aging Among Older Adults
The literature examining how to shift middle-aged and older adults’ views of aging is still relatively small. Most commonly, researchers have successfully shifted middle-aged and older adults’ views of aging through education components of physical activity interventions. Research teams created curriculum addressing views of aging, ranging from one 20-min unit (Wolff et al., 2014) to four weekly units (Beyer et al., 2019; Sarkisian et al., 2007). These interventions targeted and measured different aspects of views of aging (age stereotypes vs. perceptions of own aging) and consistently improved at least some age stereotypes (e.g., perceptions of older adults as satisfied and optimistic; Wolff et al., 2014) or even multiple aspects of views of aging (e.g., subjective age, age stereotypes, awareness of age-related change; Brothers & Diehl, 2017). Moreover, participants whose views of aging became more favorable had greater increases in physical activity in some studies (Sarkisian et al., 2007; Wolff et al., 2014). These were multicomponent interventions, such as addressing negative age stereotypes, promoting positive views of aging, and discouraging attribution of sedentary behavior to age (Sarkisian et al., 2007). Additional research is necessary to identify the essential components and whether simplified interventions could still be effective.
Another intensive intervention implemented four weekly sessions of implicit positive age stereotype priming to increase positive views of aging and improve physical function (Levy et al., 2014). The researchers also tested an explicit priming condition (writing essays about a healthy older adult) which strengthened positive views of older adults but did not change self-perceptions of aging or physical function. These interventions described thus far were time-consuming and resource-intensive. Helpful for targeting smaller (e.g., high risk) groups, they may not be as feasible to implement on a large scale. Promising evidence indicates even brief, simple interventions can have a positive effect on views of aging (e.g., a savoring writing exercise, Smith & Bryant, 2019). With evidence that views of aging are modifiable, a next important step is to identify scalable public health interventions such as public messaging.
A recent online experiment indicates that brief anti-ageism messages can shift views of aging (Busso et al., 2019). The Reframing Aging initiative developed framed messages to reduce negative attitudes toward older adults and change public discourse to improve advocacy for issues affecting us all as we age. Initial research from the project summarized the disconnect between views of gerontology experts and views of a typical adult below age 60 (Lindland et al., 2015); the initiative appears to focus most on changing younger and middle-aged adults’ attitudes. Even so, Busso and colleagues’ (2019) online study used a representative sample of Americans aged 18-92 years old to show brief public messaging can help reduce negative attitudes toward older adults. Participants who were randomly assigned to read the framed anti-ageism messages about: (a) building momentum with age, (b) implicit age bias and its consequences, and (c) benefits of intergenerational community centers showed less bias against older adults in an Implicit Attitude Test (IAT) compared with participants who read a statement unrelated to aging. The researchers reported comparable IAT scores among participants below and more than age 55 in the control condition, but they did not report whether age was tested as a moderator of the framing effect.
Greater knowledge of aging is linked to less ageism across the lifespan (Cooney et al., 2020), but reading the same anti-ageism message may not have the same effect on younger versus older adults. Considering the specific audience is crucial. For example, the Reframing Aging initiative argues individualism (i.e., focus on personal responsibility and control) can undermine support among adults below age 60 for social programs supporting older adults (Lindland et al., 2015). Conversely health messaging for older adults commonly emphasizes personal control in order to encourage older adults to enact health behaviors (e.g., Sarkisian et al., 2007). Therefore, we designed a brief anti-ageism message intervention specifically for independent adults in mid and later life to test whether simple, public messaging could encourage health behaviors at a busy senior center.
Current Study
When generating the brief anti-ageism messages, we took into account the multiplicity in views of aging and ap-proaches to shifting them in public messaging. Views of aging are multidimensional, people can be aware of both gains and losses (e.g., Diehl & Wahl, 2010) and can have different attitudes and beliefs across domains of aging such as physical versus social changes (Kornadt & Rothermund, 2014). Therefore, reducing negative views of aging and promoting positive views of aging can in turn be independent approaches with potentially different effects. Some public messaging strategies address and debunk myths and negative views of aging (e.g., American Association of Retired Persons’s [AARP] #DisruptAging campaign), while others promote positive images of aging in advertisements for products and services targeting older adults. However, these approaches have weaknesses as well. Cognitive psychology suggests that presenting myths and then retracting them can unintentionally reinforce the information through exposure (Lewandowsky et al., 2012). Promoting positive views of aging may backfire if aging realities do not live up to high expectations, and older adults can reject extremely positive images of aging as unrealistic (Fung et al., 2015). Another strategy is to discourage thinking about older adults as a homogeneous group; highlighting individuating characteristics and diversity inherent in aging may help discourage fatalism and reliance on age stereotypes.
This project intervention disseminated anti-ageism messages designed to: (a) reduce negative views of aging, (b) promote positive views of aging, or (c) emphasize diversity in aging as a way to reduce stereotype reliance. We assessed whether any brief message addressing views of aging increased motivation to engage in health promotion (e.g., to attend physically active programming) at senior centers compared with a control group, and we tested for differences in efficacy across the anti-ageism intervention framings. We expected that these messages would increase motivation to engage in health-promoting activities by shifting self-perceptions of aging and perceived control over aging outcomes. To evaluate whether the effect was specific to physical activities (i.e., activities especially salient for health promotion), we also examined whether the messages increased interest in other center activities.
Moreover, we tested whether the anti-ageism intervention effects differed based on participant age. Participants closer to 50 who identify as middle-aged may not find the messages as self-relevant as older participants. Conversely, the oldest participants could find the messages less persuasive if they rely on their own past experiences with aging more than younger participants do.
Method
Participants
The research team recruited 349 participants from community centers serving older adults (Reduce Negative: n = 84, Promote Positive: n = 89, Emphasize Diversity: n = 86; Control: n = 90); 176 from four sites in the greater Los Angeles area and 173 from three sites in the Chicago area. The sample ranged from 50 to 92 years old with a mean and median age of 72 (SD = 9, n = 331) years, was 80% female, and was racially/ethnically and socio-demographically diverse (43% Black/African American, 36% Non-Latinx White, 10% Latinx, 12% other race/ethnicity; 43% completed Bachelor’s degree). Most participants were long-time center attendees; 79% had been coming to the center for more than a year, and 69% attended more than once a week. Characteristics of participants separated by condition random assignment are displayed in Table 1.
Participant Self-Reported Characteristics, Separated by Condition.
Note. The only characteristic that differed statistically between the control and combined intervention conditions was motivation to attend physically active programs (p = .012).
Materials and Procedure
Anti-ageism messages (intervention conditions)
The anti-ageism messages created for the study each discussed physical, cognitive, and social/emotional aging and were matched for length and reading level (see Supplemental Appendix A for full text). The Reduce Negative message presented common negative age stereotypes alongside corrections to these exaggerated or false beliefs. The Promote Positive message did not mention negative beliefs and only described positive growth or maintenance in function with age. Finally the Emphasize Diversity message highlighted the complexity of aging across domains (e.g., describing both gains and losses with aging, how people age differently). To encourage attention to the messages, participants were instructed to underline sentences they found most important and to rate on 5-point scales how much they liked the message, agreed with it, and thought the information would be helpful for more people to know.
Senior center programming attendance motivation
Participants rated on a 5-point scale how likely they were to attend 16 different senior center programs if they were offered (1 = not at all likely, 5 = extremely likely). Five of these activities were physically active: yoga, dancing, walking club, exercise class, and tai chi (α = .74). Other programs included bridge/strategic card games, art class, mahjong, book club, movie screening, computer class, current events discussion group, writing group, meals, music performances by guest musicians, and guided meditation (α = .81). When participants rated at least 80% of the items, mean composite scores were generated for physically active programs (n = 341) and for other programs (n = 338).
Views on aging
Participants reported self-perceptions of aging using the AgeCog Physical Losses and Ongoing Development subscales (Wurm et al., 2007; αs = .83 and .53), rating statements such as “Aging means to me that I am less energetic and fit” and “Aging means to me that my capabilities are increasing” (1 = definitely false, 4 = definitely true). Participants also responded to three face-valid items (α = .63) rating personal perceived control over aging outcomes in physical, cognitive, and social domains such as “There are actions that I can take to maintain or improve my physical health as I grow older” or “I can take steps to maintain strong relationships throughout my life” (1 = strongly disagree, 7 = strongly agree).
Demographics
Participants reported age, gender, race/ethnicity, education, how long they had been attending the center, and how often they come.
Procedure
Researchers recruited adults aged 50 years old or older who could complete a written survey by advertising an opportunity to complete a 15-min questionnaire on “views of aging and wellness” in senior center common areas (e.g., lobbies or meeting spaces). Participants were told that the researchers would ask about three different topics: (a) how people think about aging, (b) interest in community center activities, and (c) motivation to engage in health-promoting behaviors (like attending exercise classes). The questionnaire packet’s cover page described it as a “set of questionnaires” and each questionnaire had its own title and instructions. Thus the measures appeared separate to reduce demand effects.
After participants provided written informed consent, they were randomly assigned to one of the three anti-ageism intervention conditions or the control condition. A random-number generator was used to predetermine the random order that the surveys would be administered at each senior center, and the surveys were sorted in that order before the data collection session. During data collection each participant received the next survey in the pile. Packets for the intervention conditions began with the assigned anti-ageism message, followed by the senior center programming attendance motivation, views of aging, and finally demographics measures. For the control participants, the anti-ageism message appeared at the end of the survey with the demographics measures. Control participants were asked to rate an anti-ageism message (counterbalanced) so that all participants’ packets were of equal length. There was no indication of the experimental condition on the cover of the survey; therefore, researchers were blind to condition during data collection. Participants received a US$20 incentive for their time. The procedures and materials were approved by the University of California—Los Angeles and Mather Institutional Review Boards.
Data Analysis
To account for clustering of responses within senior centers, we tested multilevel models using the Stata 14 mixed command. We used an intention-to-treat approach to compare motivation to attend physically active programs (or other programs) in the control and anti-ageism intervention conditions. We first collapsed the intervention conditions and then looked at the statements separately.
In each set of analyses we tested whether age moderated the relationship between the condition and outcome with age centered at the mean of 72 years old. If the interaction term was significant, we determined the regions of significance (i.e., the age range where the control and intervention group differ significantly) using Preacher et al. (2006) online tool for hierarchical linear models and presented observed means to demonstrate the differences. When the model included multiple interaction terms, we used the Stata contrast command to test the overall significance of age as a moderator and plotted marginal effects and recentered predictors to describe significant interactions. When comparing experimental groups, we used a Bonferroni correction based on the number of comparisons within the model.
Results
Differences in Senior Center Programming Attendance Motivation
Physically active programming
The anti-ageism intervention effect differed for younger and older participants; interaction b = −0.04, SE = 0.01, p = .004, n = 325 (Figure 1). 1 Among participants over age 71 (54% of sample), participants in an intervention condition were more interested in attending physically active programming (M = 3.35, SD = 0.94) than those in the control group (M = 2.87, SD = 0.99). Among participants aged 71 and younger, there was no equivalent difference; intervention M = 3.44, SD = 0.88 versus control M = 3.45, SD = 0.82. Interestingly, within the control group, older participants were less interested in physically active programs than younger participants (age b = −0.05, SE = 0.01, p < .001), but there was no comparable age difference in the intervention groups (b = −0.01, SE = 0.01, p = .174).
Average motivation to attend physically active programs at community center. Shaded area represents 95% confidence intervals.
When differentiating among anti-ageism intervention statement types, the effect of condition on physical activity interest again differed by age, contrast χ2(3) = 9.95, p = .019. Investigation of simple effects showed that although older participants in the control group were less interested in physical activities than younger participants (b = −0.05, SE = 0.01, p < .001), there were no comparable age differences in interest in physical activities among participants in any of the experimental conditions (ps ≥ .10). Pairwise comparisons between the experimental conditions revealed no evidence of differences between the intervention groups’ interest in physical activities (ps ≥ .24 when age centered at 10th, 50th, and 90th percentiles).
Other programming
The anti-ageism intervention effect on participants’ interest in other programs also differed by age; interaction b = −0.03, SE = 0.01, p = .009, n = 324. Among participants aged 64 and below (20% of sample), those assigned to read an anti-ageism message had lower interest in the less physically active programs (M = 3.32, SD = 0.71) than peers in the control condition (M = 3.51, SD = 0.59). Conversely, among participants aged 84 and older (10% of sample), those assigned to read an anti-ageism message had higher interest in the other programs (M = 3.20, SD = 0.80) than peers in the control group (M = 2.78, SD = 0.45). Among participants aged 65-83 years old (70% of sample), those in the control (M = 3.16, SD = 0.83) and intervention groups (M = 3.13, SD = 0.73) did not differ from each other. Again, within the control group older participants were less interested in other programming than younger participants (b = −0.03, SE = 0.01, p = .004), but there was no age difference in interest overall across the intervention groups (b = 0.001, SE = 0.01, p = .82).
Specifically, differences between the control and anti-ageism intervention groups appear to be driven by the Promote Positive group. Age moderated the effect of specific condition on interest in other activities, contrast χ2(3) = 9.98, p = .019. Plotting marginal effects indicates that only the Promote Positive condition differed significantly from the control group, specifically among the younger participants (Figure 2). For example, at age 61, only Promote Positive participants had lower predicted interest than the control condition (b = −0.66, SE = 0.19, p = .001). The Reduce Negative group had 0.48 higher predicted interest in other activities than the Promote Positive group (SE = 0.19, p = .011) and the Emphasize Diversity group also tended to have higher predicted interest (b = 0.41, SE = 0.18, p = .020).
Average difference between intervention and control conditions in motivation to attend other activities. Shaded areas represent 95% confidence intervals. The promoting positive condition 95% confidence interval does not include 0 for younger participants’ average difference from the control condition, indicating they were less motivated to attend other activities than peers in the control condition.
Views on Aging as Candidate Mediator
The views on aging measures were not normally distributed. AgeCog Physical Losses composite scores appeared bimodal; 1.5 (between definitely false and somewhat false) and 3 (somewhat true) were the most common scores. Most participants agreed strongly with the AgeCog Ongoing Development statements (median = 3.5 on 4-point scale). Similarly most participants agreed strongly with statements indicating perceived control over aging outcomes (median = 6.3 on 7-point scale). Kruskal–Wallis tests (nonparametric tests of the hypothesis that several samples are from the same population) did not show evidence that the control and intervention conditions differed in any of the measured views of aging, even specifically among participants over age 71, ps > .09. The lack of an association between the independent variable (intervention vs. control) and views on aging is not consistent with views on aging being a mediator of the intervention effect (Baron & Kenny, 1986).
Notably, AgeCog Physical Losses (b = −0.23, SE = 0.06, p < .001) and perceived control over aging outcomes (b = 0.18, SE = 0.06, p = .004) were independently associated with motivation to attend physical activities. 2 The intervention effect on motivation to attend physically active programs was significant over and above the associations between these views of aging and motivation; for example, at age 72, the control group’s interest was 0.22 lower than the intervention group’s; SE = 0.11, p = .040. 3
Discussion
As hypothesized, senior center attendees who were randomly assigned to read anti-ageism messages were more motivated to attend physically active programs than control participants. However, the effect differed by age; specifically, senior center attendees aged 72 years old and older in an anti-ageism intervention condition were more motivated to attend physically active programming than peers assigned to not read a message. Younger participants may not have felt like an “older adult” yet and the messages may not have been as self-relevant (similar to other studies where only late middle-aged adults who had older self-identity showed self-stereotyping, O’Brien & Hummert, 2006). Thus, anti-ageism messages may have a bigger impact on older participants’ attitudes because the messages could affect self-perceptions of aging as well as views of other older adults. Future longitudinal research could also provide insight on whether individuals’ perceptions of anti-ageism messaging change as they age.
Participants assigned to the different anti-ageism message versions did not differ in their interest in physically active programming. Although we created new anti-ageism messages for this study to manipulate emphasis on reducing negative views of aging versus promoting positive views of aging, the lack of observed differences across framings suggests some flexibility in the anti-ageism message content. Therefore, future research should test whether anti-ageism messages from the Reframing Aging project (Busso et al., 2019) also promote motivation to engage in physical activities among older adults given that these messages are already being actively, widely disseminated as part of the funded initiative. Moving beyond motivation, additional research is needed to examine the effect of anti-ageism messages on levels of physical activity and program attendance.
We evaluated the anti-ageism intervention impact on both physical activities and other programming at the senior centers to see whether the effect was specific to physical activity promotion. The positive effect of reading an anti-ageism message on older participants’ motivation to attend activities was strongest for physically active programs. This may be due to the strong association between physical activity and health promotion; physical activity may be an investment in one’s future, whereas other center activities may be seen as primarily for current pleasure. Future research could investigate whether anti-ageism messages more effectively promote interest in other activities when they are explicitly framed as promoting cognitive function and long-term well-being.
Surprisingly, we also observed younger participants assigned to read the message promoting positive views of aging were less interested in other programming than peers in the control condition. Notably, very few control participants were below age 65 (5% of the sample; 18% of the control group), and younger participants may have differed from the others in more substantial ways such as being caregivers or volunteers. Historically, the average age of senior center attendees has been 75 years old (National Council on Aging, 2012); however, newer community center models aim to engage younger older adults (Pardasani & Thompson, 2012). To better understand whether promoting positive views of aging can backfire among younger senior center attendees, future work could oversample attendees below age 65 to see if the effect replicates.
Future messaging efforts might also consider how different messages might appeal to attendees based on the duration or frequency of their participation in programming at the center, considering that these results are based on a significant number of attendees coming to the center for more than 1 year (79%) and who attend more than once per week (69%). Newer or less frequent attendees may naturally be exposed to less messaging overall, potentially weakening the impact of such messaging. However, these attendees may also have greater room for improvement in motivation to attend activities.
Previous work has linked self-perceptions of aging to preventive health behaviors (e.g., Levy & Myers, 2004) and improvements in different views on aging to increased physical activity (e.g., Sarkisian et al., 2007; Wolff et al., 2014), but we were unable to show that differences in views on aging explained the effect of this intervention on interest in physical activities. Lack of difference in views on aging between the control and intervention groups may be due to measure psychometrics (e.g., scale reliability and distribution challenges). Perhaps due to self-selection among older adults attending senior centers, even the control condition participants already had very positive views of aging and strongly endorsed beliefs that they could control aging outcomes through personal behaviors.
Moreover, the increased motivation to attend physically active programs could simply not be explained by differences in these views on aging. Aging identity and behaviors may be impacted by a variety of views of aging including perceptions of how others treat older adults and own attitudes toward other older adults (Gendron et al., 2020). Our measures of self-perceptions of aging only captured a subset of views of aging. We also cannot rule out the possibility that the increase was due to priming the concept of physical activity, and future work should include a stronger control condition to test whether the age-framing specifically affected motivation. Regardless of the mechanism, given participants in their 80s and older who read these brief anti-ageism messages rated themselves on average as almost a full point more likely to attend physically active programs (on a 5-point scale) than their peers in the control group suggests this intervention merits further consideration.
It is notable that we observed this effect even in a busy senior center environment with distractions from ongoing activities. Our high participant incentive also encouraged diverse participation, not just from center attendees who were especially interested in research. Therefore, although our results are specific to attendees of senior-focused community centers in greater metropolitan areas, the findings are ecologically valid within this population of senior center attendees. Additional research is needed to assess the duration of the observed effect; however, even a brief effect may be useful if anti-ageism messages promoting adaptive views of aging appear alongside senior center activity catalogs or near activity sign-up opportunities. There are many opportunities at senior centers for repeated exposure to these kinds of messages.
Based on the current findings, incorporating anti-ageism messages into promotions for physical activity programming has strong potential for encouraging program attendance. Although the study focused on promoting interest in programs among senior center attendees, these findings could extend to senior living communities and other aging services as well. Future research could even investigate whether including anti-ageism messages along with information about online physical activity offerings when older adults cannot leave their homes promotes activity engagement. Messages addressing views of aging have the potential to be a low-cost, wide-reaching means of promoting healthy behaviors in older adults.
Supplemental Material
sj-pdf-1-jag-10.1177_0733464820960925 – Supplemental material for Brief Anti-Ageism Messaging Effects on Physical Activity Motivation Among Older Adults
Supplemental material, sj-pdf-1-jag-10.1177_0733464820960925 for Brief Anti-Ageism Messaging Effects on Physical Activity Motivation Among Older Adults by Josephine A. Menkin, Jennifer L. Smith and Joseph G. Bihary in Journal of Applied Gerontology
Footnotes
Acknowledgements
The authors acknowledge and thank Dr. Theodore Robles, Dr. Catherine Sarkisian, and Dr. Teresa Seeman for their support consulting on project preparation and data collection, as well as Dugan O’Connor, Alyssa Choi, Jessica Morales, Lasya Gudipudi, and Maribel Garcia for their assistance with data collection. They also thank the staff and participants at the community center sites for welcoming and sharing their thoughts. During the time of this research, Dr. Josephine Menkin’s affiliation was with the Division of Geriatrics, Department of Medicine, David Geffen School of Medicine at UCLA. The views expressed in this publication are the author’s own and do not purport to reflect the views of the author’s employer.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by a collaborative research grant from the Mather Institute [award number 20190310].
Ethical Approval
Research was approved by the University of California, Los Angeles (UCLA) South General IRB (Protocol #18-001887) and the Mather IRB (Protocol #18-001).
Supplemental Material
Supplemental material for this article is available online.
Notes
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
