Abstract
Quality of life outcomes are associated with the transition to non-driving and depend on effective coping. We examined the relationship between internal factors associated with effective coping and longitudinal changes in travel behavior among drivers aged ≥70 using data from a randomized controlled trial evaluating a driving decision aid. We measured attitudes using the Assessment of Readiness for Mobility Transition (ARMT) and personality using the Ten-Item Personality Inventory (TIPI). We analyzed associations between attitudes and personality with drivers’ change in a) alternative transportation use and b) driving behavior over time. Older drivers with high versus low attitudinal readiness for mobility transition were more likely to use alternative transportation over time (adjusted odds ratio = 6.37; 95% confidence interval: 1.45–28.1). We found no association between personality characteristics and alternative transportation use or driving behavior over time. Attitudinal readiness may be a key predictor of effective coping during the transition to non-driving.
• Older drivers with higher levels of attitudinal readiness for a mobility transition have greater odds of using alternative transportation options to driving compared to those with low readiness. • Our findings provide additional evidence to support the conceptual basis of the Assessment of Readiness for Mobility Transition that individuals with high readiness would be more open to various pathways to achieving their mobility.
• Validated tools to identify internal and external factors that influence effective coping may be particularly useful for clinicians to help prompt conversations about planning for the transition to non-driving and guide recommendations based on patients’ responses. • Increasing readiness for a mobility transition may help older drivers to be more open to a variety of transportation options to meet their mobility needs as they reduce or stop driving.What this paper adds?
Applications of study findings
Introduction
As people age, changes in health and function may necessitate changes in driving or retirement from driving (Edwards et al., 2009). Decisions about limiting or stopping driving are often emotionally and logistically difficult for older adults, their family, and friends (Fowler et al., 2023). For clinicians, decisions about patient driving can be complicated by a lack of knowledge about how to best support patients faced with these decisions and, in some states, mandatory reporting to authorities if the patient has specific health conditions. Driving is the primary mode of away-from-home travel for older adults in the United States (Lidbe et al., 2021; Shen et al., 2017) and has strong ties to independence and identity (Sanford et al., 2020). Older adults report experiencing negative emotions when discussing driving cessation, including sadness, frustration, and powerlessness (Betz et al., 2016). Such feelings are often matched by complex interpersonal and logistical scenarios for families (Lafrance et al., 2022; Veerhuis et al., 2023). It is important for clinicians to support older adults and their families in decisions about driving, including patient’s emotional wellbeing and their out-of-home mobility.
How individuals and their families cope with the transition to non-driving has implications for quality of life (Choi, Adams, & Mezuk, 2012). Coping with a life transition such as driving cessation typically depends on antecedents such as triggering events about safety (e.g., a car crash or near miss) and the individual’s or family’s appraisal of the seriousness of these triggering events. The processes of appraising and coping are further influenced by patients’ external factors (e.g., social and environmental context) and their internal factors (e.g., attitudes and personality characteristics) (DeLongis & Holtzman, 2005). Previously studied external factors have shown predictable associations between potential coping strategies and the ability to transition to non-driving. Older drivers who receive transportation support from their social network have higher odds of reducing or stopping driving (Choi, Adams, & Kahana, 2012; Hansmann et al., 2024; Ichikawa et al., 2016). These findings support a conceptual model of mobility that is made up of different types of capital (i.e., infrastructure, social, cultural, and individual) one can leverage to enable mobility even after no longer driving (Musselwhite & Scott, 2019). However, the relationship between internal factors (e.g., attitudinal and dispositional characteristics) and transitions to non-driving may be more complex. For example, older drivers with higher levels of extraversion and resilience have lower odds of reducing or stopping driving (St Louis et al., 2020; Stinchcombe et al., 2023). Internal factors that could represent greater individual capital are not necessarily associated with progressing through transitions to non-driving. Attitudes and personality characteristics may play a role in older drivers’ adaptability and preparedness to use alternative forms of transportation but may also be associated with longer periods of independent driving. A better understanding of internal factors that influence effective coping is needed to best support older drivers, their family, and friends through these transitions.
Developing effective supports for the transition to non-driving is further complicated by the fact that many older adults are unaware of a need to plan until a crisis arises (Kandasamy et al., 2018). Understanding the impact of internal factors associated with better coping in the transition to non-driving can inform effective strategies to tailor existing resources and interventions. Examples include detailed websites, online decision support tools, and coaching programs (Dickerson et al., 2024) to help older adults and their care partners develop effective coping strategies to navigate this common mobility transition.
In this study, we use data from the Advancing Understanding of Transportation Options (AUTO) study (Betz et al., 2021) to characterize internal factors (i.e., readiness for a mobility transition and personality characteristics) that may influence coping strategies used in transitions to non-driving. We focused on two outcomes for predictive modeling: (1) reported usage of non-driving alternative transportation; and (2) reported reductions in driving mobility. We hypothesized that higher attitudinal readiness for a mobility transition would be associated with greater odds of transitioning to non-driving at baseline and over time as measured by using alternative transportation options and reducing driving behavior. We also hypothesized that personality characteristics including greater openness to new experiences would be associated with greater odds of transitioning to non-driving at baseline and over time.
Methods
Study Population
AUTO is a randomized trial of a decision aid to support older adults and their chosen partners (e.g., family member or friend, referred to hereafter as “study partners”) about driving decisions with follow-up at 6, 12, 18, and 24 months. The AUTO study protocol and intervention is described in detail elsewhere (Betz et al., 2021); briefly, participants were eligible if they were 70 years old or older, spoke English, drove at least once per week, had a valid driver’s license, had a primary care provider at one of the three study sites, and had at least one medical condition associated with reduced driving ability (e.g., glaucoma, dementia, and seizure). Investigators excluded otherwise eligible individuals if cognitive screening suggested moderate to severe dementia, as indicated with a 5-min Montreal Cognitive Assessment (MoCA) score <21 (Wong et al., 2015). Study participants were recruited between December 2019 and June 2021 and completed follow-up assessments at six-month intervals across 24 months.
Recruited eligible and interested older drivers provided informed consent and were asked to identify a family member or friend (i.e., someone who might be involved in the decision-making process about driving or provides support for the transition to non-driving) as a study partner, who was dually enrolled to assess their influences on participant driving decisions and outcomes. The parent study was approved by the institutional review boards of study sites (University of California San Diego, University of Colorado, and Indiana University). The trial is registered with clinicaltrials.gov (Clinical Trials.gov Identifier NCT04141891). In this exploratory analysis, we included relevant variables regarding study enrollment procedures in models including the study site and whether the participant enrolled before or after the onset of the COVID-19 pandemic.
Study Measures
We targeted several internal factors that influence coping as potential predictors in this study. Internal factors associated with the potential for effective coping strategies included readiness for transition (as measured by the Attitudinal Readiness for Mobility Transition [ARMT] (Meuser et al., 2013); and five personality characteristics (extraversion, agreeableness, conscientiousness, emotional stability, and openness to experiences) as measured by the Ten-Item Personality Inventory (TIPI) (Gosling et al., 2003). The ARMT tool (available online at https://www.umsl.edu/mtci/PDFs/ARMT_2011c.pdf) includes 24 statements that participants rate how strongly they agree or disagree on a scale of 1–5 across domains related to driver perception. The ARMT tool includes constructs of anticipatory anxiety (e.g., “asking others for help with mobility means that I am losing my independence”), perceived burden (e.g., “I am a burden if I ask others for help with transportation”), avoidance (e.g., “I avoid thinking about losing my mobility”), and perceived adverse situations (e.g., “mobility loss can be sudden or progressive, but it is always devastating”) (Meuser et al., 2013).
Forty-six participants skipped at least one ARMT statement, and, of those, they skipped an average of 1.8 statements (standard deviation 1.3), with the most common skipped statement during interviews being “I wish others would stop talking to me about my mobility” (n = 36). For this study, missingness was imputed by averaging individual total ARMT response scores (Meuser et al., 2013). We categorized participants’ ARMT mean scores into three groups for clinical interpretability: low readiness (ARMT >3.57), mixed readiness (ARMT 2.29–3.57), and high readiness (ARMT <2.29) (Meuser et al., 2013). Individuals with high readiness are accepting of the uncertainties of aging, they understand that some decline is likely, and they are open to various pathways to achieve their mobility and other functional aims. Drivers with mixed readiness are likely to have mixed emotions if they were to experience a sudden change in their ability to drive independently. Finally, those individuals with low readiness might struggle to adapt if they experience a sudden change in their mobility.
The TIPI is made up of 10 questions using a 1–7 scale to assess the five personality characteristics. Scores range from 1 to 7 with higher scores indicating greater presence of that personality characteristic (Gosling et al., 2003).
We included two self-reported primary outcomes of interest: use of alternative transportation in the past three months and any driving reduction in the past six months. We defined use of alternative transportation as self-reported use in the past three months of public bus, local tram/train/subway, rideshare service (e.g., Lyft or Uber), taxi, special community transportation (e.g., Dial-A-Ride), ride with a friend/family member as a passenger, ride with another volunteer driver as a passenger, and/or other modes of transportation (e.g., bicycling/e-biking). We defined any driving reduction based on self-reported reductions in driving behavior in the past six months including: reduced trips driven, reduced days driven, reduced miles driven, or reduced trip distance. We treated these individual driving behavior changes as secondary outcomes of interest.
Additional Variables
We assessed self-reported demographic characteristics of this sample including age, race ethnicity, education as the highest grade completed, current living arrangement, and neighborhood Area Deprivation Index (ADI) national percentile (continuous). ADI scores are calculated using census data and is a composite measure of neighborhood social vulnerability (Kind & Buckingham, 2018). ADI national percentiles range from 1 to 100 with higher percentiles representing more disadvantaged neighborhoods.
We included measures of the participants’ cognition using the 5-min Montreal Cognitive Assessment (MoCA) tool using accepted cut points (Wong et al., 2015). We also included Patient-Reported Outcomes Measurement Information System (PROMIS) measures of physical and mental health (HealthMeasures, 2020).
Analysis
We summarized predictors, covariates, and other relevant demographic variables at baseline with frequencies and percentages for categorical variables and with means and standard deviations (SD) for continuous variables. We calculated the rates of the transportation use and driving reduction outcomes at the baseline interview and over time for our sample. We determined initial model covariates a priori for their associations with driving/transportation outcomes (baseline age, gender, Area Deprivation Index (ADI), PROMIS mental health T-score, and PROMIS physical health T-score), as well as structural covariates (site, whether the baseline visit was during or pre-COVID).
We tested whether other demographic attributes were associated with any coping factors to determine whether they should be included as additional confounders (not shown), finding that whether the participant lived alone was often highly informative and deserved inclusion in all models as a confounder. For longitudinal outcome models, we also included the randomization group and study time point as a categorical variable to allow for non-linear time trends. Finally, since we imputed ARMT scores for 15.3% of participants who were missing a response to at least one of the 24 ARMT statements, in all models with ARMT as a predictor, we included an indicator for whether the ARMT score was imputed, thereby adjusting for differences between individuals with complete and partially missing surveys. The assumption underlying this approach is that the survey responses are missing at random.
We fit two sets of models. First, cross-sectional logistic regression models with baseline predictors and baseline outcomes, to determine whether each predictor was associated with each outcome at baseline. Second, longitudinal generalized linear mixed-effects models with baseline predictors and longitudinal outcomes (6-, 12-, 18-, and 24-month follow-up visits), to determine whether these associations changed or persisted across time. For baseline models, we fit each baseline predictor versus baseline outcome combination in a separate logistic regression model with the covariates previously listed. For longitudinal models, we fit each baseline predictor versus longitudinal outcome combination in a separate generalized linear mixed model with binomial distribution and random intercept for each participant. We conducted analyses for secondary outcomes—individual driving behavior changes—as well. Given the exploratory nature of these analyses, we did not explicitly adjust our results to account for multiple comparisons (five predictors and two outcomes). We performed all statistical analyses in R version 4.3.1 (R Core Team, 2023) and fit longitudinal models using the lme4 package (Bates et al., 2015).
Results
Driver Demographics, Study Enrollment, Health Status, and Internal Factors That Influence Coping for Participants (n = 301) at Baseline.
Abbreviations: SD: standard deviation. PROMIS: Patient-Reported Outcomes Measurement Information System. TIPI: Ten-Item Personality Inventory.
aMore than one response is allowed, so percentages do not sum to 100%.
At enrollment, most participants (65.8%) had mixed readiness for a mobility transition based on their ARMT scores, yet, one-tenth of the sample (10.6%) had high readiness for a mobility transition and nearly one in four participants (23.6%) had low readiness for a mobility transition. Participants’ mean (SD) Ten-Item Personality Index scale scores were as follows: emotional stability [5.8 (1.1)], extraversion [4.5 (1.6)], agreeableness [5.7 (1.1)], conscientiousness [5.9 (1.0)], and openness to experiences [5.3 (1.2)]. These generally reflect above average representation of these characteristics compared to previously reported norms (Gosling et al., 2014).
At baseline, most participants reported having used an alternative mode of transportation in the past three months, with the most common mode being a ride from a friend or family member (83.1%, Appendix Table 1). More than half the participants had reduced their driving in some way over the six months before enrollment (61.0%, Appendix Table 1). When analyzing associations between internal factors that influence coping with the odds of outcomes of interest at baseline, we found no association between internal factors, such as attitudinal readiness for a mobility transition, and the odds of alternative transportation use in the three months prior to the baseline evaluation in adjusted models (Appendix Table 2).
Odds of Alternative Transportation Use Over Study Follow-up Visits Associated With Internal Factors That Influence Coping.
Footnote: Each longitudinal binary outcome is modeled in a generalized linear mixed-effect model with a binomial distribution. Adjusted models include driver age, gender, Area Deprivation Index (ADI), whether the baseline visit was during or before COVID, time point, study randomization group, site, PROMIS physical and mental health T-scores at baseline, and whether the driver lives alone. Adjusted ARMT models include an indicator for whether participants had an imputed ARMT score or had answered all ARMT questions. All models included a random intercept for each participant to account for repeated measures over time.
Odds of Reducing Driving Over Study Follow-up Visits Associated With Internal Factors That Influence Coping.
Footnote: Each longitudinal binary outcome is modeled in a generalized linear mixed-effect model with a binomial distribution. Adjusted models include driver age, gender, Area Deprivation Index (ADI), whether the baseline visit was during or before COVID, time point, study randomization group, site, PROMIS physical and mental health T-scores at baseline, and whether the driver lives alone. Adjusted ARMT models include an indicator for whether participants had an imputed ARMT score or had answered all ARMT questions. All models included a random intercept for each participant to account for repeated measures over time.
Discussion
Decisions to change driving habits or to stop driving are common as people age; however, a full understanding of what contributes to better coping during and after this transition is a knowledge gap in supporting older drivers. In this exploratory analysis, we evaluated whether internal factors that influence coping such as attitudinal readiness for transitioning to non-driving and personality characteristics of older primary care patients were associated with the odds of using alternative transportation options and the odds of reducing driving behavior over time. We found that older drivers with higher readiness for a mobility transition at baseline had greater odds of using alternative transportation options over the study period. These findings can guide future confirmatory research to investigate both internal and external factors that influence coping that could be used to tailor interventions and policies to the different needs of older drivers during the transition to non-driving.
Previous research investigating characteristics of older drivers based on their readiness for a mobility transition found that older drivers reported no significant difference in their limited ability to access alternative transportation options regardless of their readiness (Lee & Scott, 2022). Another study investigating readiness for a mobility transition characterized people with low readiness as being less ready to transition to non-driving based on feelings of anxiety about loss of personal independence and worry about becoming a burden to others (Meuser et al., 2013). Despite the prior evidence suggesting no significant differences in access to alternative transportation options, our findings support an association between higher readiness for a mobility transition and higher odds of using alternative transportation options across time. This is consistent with the features of readiness assessed using the ARMT—namely, that individuals identified as having high readiness according to the ARMT are accepting of the functional declines associated with aging and open to using a variety of options to meet their mobility goals (Meuser et al., 2013). Further research is needed to characterize the relationship between attitudinal readiness, use of alternative transportation options, and interpersonal outcomes such as sense of personal independence and self-sufficiency over time.
We found no association between readiness for a mobility transition and driving reduction over 24 months of follow-up in this study. However, the association between higher readiness for a mobility transition and greater odds of using alternative transportation options over time suggests it may be important to observe longer follow-up periods to determine whether readiness for a mobility transition may be associated with driving reduction. In addition to conducting studies with longer follow-up periods, future research can gain insights about how readiness for a mobility transition influences the transition to non-driving by investigating the association between readiness and decision quality, psychosocial, and mobility outcomes for older drivers as they reduce driving behavior and ultimately stop driving. Given that driving cessation can cause emotional distress (Sanford et al., 2020), it is important to measure how readiness for a mobility transition may be associated with outcomes such as decision conflict (AM O’Connor, 2006), depression, regret, and life space constriction (Baker et al., 2003) after driving reduction. Future studies of readiness for a mobility transition may find that those with high readiness are less likely to experience these downstream adverse outcomes compared to those with low readiness. It will also be important in future work to assess the extent to which current driving cessation resources and interventions support quality of life and other wellbeing outcomes for older adults during the transition to non-driving (Dickerson et al., 2024).
Limitations
The AUTO study population consists primarily of older adults living in urban and suburban areas who identify as White, Non-Hispanic, and have high levels of education. This may limit the generalizability of these findings to older adults living in rural areas, where alternative transportation options to driving are much more limited (Kandasamy et al., 2017). Additionally, participants enrolled in this study on average lived in areas with low levels of neighborhood disadvantage, which may also reflect more resources and better access to alternative transportation options. Future studies should seek to include participants from rural and under-resourced areas and gather data regarding neighborhood walkability and the quality of alternative transportation options to investigate whether associations between internal and external factors that influence effective coping are similar across different sociodemographic groupings and to what extent internal and external factors are interrelated. For example, living in an extremely walkable area with high quality public transit and paratransit services may influence individuals’ attitudinal readiness for mobility transition.
Study enrollment overlapped the start of the COVID-19 pandemic, which may have led to differences in baseline driving behavior if participants were enrolled before or after March 2020, as well as differences in driving behavior at follow-up due to COVID-19-related activity restrictions. The COVID-19 pandemic influenced travel behavior in multiple ways including decreased service provided for public transit, increased avoidance of modes of travel that required being in a vehicle with others including rideshare services or rides from others, and decreased trips made for leisure (Dadashzadeh et al., 2022). Older adults’ concern for exposure to the COVID-19 virus may have contributed to greater use of driving oneself in a private vehicle (Gao et al., 2023). In a separate analysis of the data from this study, we found that compared to participants who enrolled before the COVID-19 pandemic, more participants who enrolled during the pandemic reported driving reductions (Betz et al., 2022).
Finally, this analysis was exploratory in nature. The small number of participants who had either low or high readiness for mobility transition limited our power to detect significant differences between participants based on these sub-groupings. Further research involving older drivers with both low and high readiness for mobility transition will contribute to our understanding of these internal factors’ role in the transition to non-driving.
Conclusions
Our findings suggest that internal readiness for a mobility transition is associated with higher odds of using alternative transportation options to driving. This is consistent with the conceptual basis of the Assessment of Readiness for Mobility Transition that individuals with high readiness would be more open to various pathways to achieving their mobility. Broadly, these exploratory results support the potential utility of the Assessment of Readiness for Mobility Transition for understanding older drivers’ change in travel behavior over time. Validated tools to identify internal and external factors that influence effective coping may be particularly useful for clinicians to help prompt conversations about planning for the transition to non-driving and guide recommendations based on patients’ responses. Development of future interventions and policies to address the needs of older drivers transitioning to non-driving will benefit from an understanding of key internal and external factors, including older adults’ readiness for mobility transition.
Supplemental Material
Supplemental Material - Internal Factors that Influence Coping in Older Drivers’ Transition to Non-Driving
Supplemental Material for Internal Factors that Influence Coping in Older Drivers’ Transition to Non-Driving by Kellia J. Hansmann, Thomas Meuser, Rachel L. Johnson, Ryan Peterson, Nicole R. Fowler, Carolyn G. DiGuiseppi, Duke Han, Ryan Moran, Faris Omegaric, and Marian E. Betz in Journal of Applied Gerontology.
Footnotes
Acknowledgments
The authors would like to thank Dr Linda Hill, Dr Christopher Knoepke, and Dr Daniel Matlock for providing suggestions and feedback as we revised the final version of this manuscript.
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 research was supported by the National Institute on Aging at the National Institutes of Health Grant Number (R01 AG059613). This project was also supported by NIH/NCATS Colorado CTSA Grant Number (UL1 TR002535).
IRB Protocol Approval Number
Colorado Multiple Institutional Review Board (COMIRB) Protocol Number: 19–0059.
Supplemental Material
Supplemental material for this article is available online.
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.
