Abstract
Background:
Older adults are faced with many unique and highly consequential decisions such as those related to finances, healthcare, and everyday functioning (e.g., driving cessation). Given the significant impact of these decisions on independence, wellbeing, and safety, an understanding of how cognitive impairment may impact decision making in older age is important.
Objective:
To examine the impact of mild cognitive impairment (MCI) on responses to a modified version of the Short Portable Assessment of Capacity for Everyday Decision making (SPACED).
Methods:
Participants were community-dwelling, actively driving older adults (N = 301; M age = 77.1 years, SD = 5.1; 69.4% with a college degree or higher; 51.2% female; 95.3% White) enrolled in the Advancing Understanding of Transportation Options (AUTO) study. A generalized linear model adjusted for age, education, sex, randomization group, cognitive assessment method, and study site was used to examine the relationship between MCI status and decision making.
Results:
MCI status was associated with poorer decision making; participants with MCI missed an average of 2.17 times more points on the SPACED than those without MCI (adjusted mean ratio: 2.17, 95% CI: 1.02, 4.61, p = 0.044).
Conclusion:
This finding supports the idea that older adults with MCI exhibit poorer decision-making abilities than cognitively normal older adults. It also suggests that older adults with MCI may exhibit poorer decision making across a wide range of decision contexts.
INTRODUCTION
Decision making is a complex process that involves the identification of a desired outcome, comparison of multiple options, and consideration of consequences. While effective decision making is important across the lifespan, older adults are faced with many unique and consequential decisions such as those related to retirement funds, allocation of resources, health care, and changes to driving or driving cessation. In addition to the inherent difficulty of these decisions, evidence suggests that common experiences that occur with age as well as brain changes that occur during pathological aging may negatively affect decision making, thereby compromising the quality of highly significant decisions [1–3]. Therefore, a greater understanding of factors that contribute to poorer decision making will help support the independence, wellbeing, and safety of older adults through the provision of interventions (e.g., decision aids, compensatory strategies, shared decision making) when appropriate.
Although it is widely accepted that decision making is impaired in older adults with dementia [2], the relationship between mild cognitive decline and decision making is less clear. Mild cognitive impairment (MCI) is a syndrome characterized by mild declines in cognition that are above and beyond normal age-related changes but do not significantly interfere with activities of daily living [4]. While prevalence estimates vary based on the method of MCI categorization, evidence suggests that approximately 16% of older adults experience mild levels of cognitive impairment [5]. Prior work has shown that individuals with MCI exhibit poorer financial and healthcare decision making [6], lower medical decision making capacity [7, 8], lower financial decision making capacity [9] and reduced capacity for consent to treatment or research participation [10–12] compared to cognitively normal older adults. However, to our knowledge no study has investigated the association between MCI and routine or everyday decision-making ability.
In this study, we examined the impact of MCI on the Short Portable Assessment of Capacity for Everyday Decision making (SPACED), which provides an estimate of participants’ general ability to engage in decision making. To investigate this question, we utilized data from the Advancing Understanding of Transportation Options (AUTO) study, a multisite study designed to investigate the use of a decision aid to assist decisions about driving in older adults [13]. We hypothesized that older adults evaluated at baseline as MCI would exhibit poorer decision making ability than older adults evaluated at baseline as cognitively normal.
MATERIALS AND METHODS
Participants
Participants in the current study were enrolled in the AUTO study, a randomized trial examining the effect of a driving decision aid among older adult drivers. Eligibility criteria and recruitment strategies have been described elsewhere [13]. Briefly, participants in the study were English-speaking older adults aged 70 or above with a valid driving license who drove at least once a week, had a primary care provider affiliated with health care systems at one of the study sites in San Diego, California; Denver, Colorado; and Indianapolis, Indiana, had been diagnosed with at least one medical condition known to be associated with impaired driving safety or increased likelihood of driving cessation, and scored 21 or above on a 5-minute Montreal Cognitive Assessment (MoCA) [14, 15] administered during telephone screening. Eligible participants were randomized into one of two groups (decision aid or control).
The Colorado Multiple Institutional Review Board (COMIRB) reviewed and approved study procedures and documents under Expedited Review, Category 7. The COMIRB approved a waiver of written documentation of informed consent and HIPAA authorization. The study protocol included a comprehensive oral consent process.
Measures
Measures used to assess cognitive status differed across participants due to a protocol change from in-person to telephone-based assessment occurring in May of 2020 in response to the COVID-19 pandemic. A study neuropsychologist (SDH) used the measures described below to categorize participants’ cognitive status as cognitively normal or MCI.
For in-person assessments of cognitive status, the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) [16] was used. The RBANS is a battery specifically designed to detect and characterize mild dementia. It takes approximately 25 minutes to administer and generates index scores for five cognitive domains. Based on prior work demonstrating good discrimination between individuals with MCI and healthy controls using the delayed memory index (DMI) scores (area under the curve of 0.90), the DMI score was used to determine participants’ cognitive status for the current study [17]. Participants with a DMI score of less than 85 were categorized as MCI.
For telephone-based assessments of cognitive status, the Brief Test of Adult Cognition by Telephone (BTACT) [18] and Oral Trail Making Test (OTMT) [19] were used. The BTACT is a battery consisting of five tests: 1) immediate and delayed word list recall (episodic verbal memory); 2) backward digit span (working memory); 3) category verbal fluency (executive function - cognitive flexibility and language); 4) number series (inductive reasoning); and 5) 30 seconds and counting task (processing speed). The OTMT is a measure of executive functioning (set shifting, working memory) and consists of two parts (A and B). In OTMT-A, participants are asked to count from 1 to 25 as quickly as possible. In OTMT-B, participants are asked to alternate between counting and letters of the alphabet as quickly as possible. Participants’ scores on each test of the BTACT and OTMT-A and -B were converted to z-scores, and participants with scores of –1 or lower on three or more tests were categorized as MCI. This type of actuarial, data-driven approach to diagnosis has been shown to be more conservative than traditional single test approaches [20], and to more reliably detect cognitive phenotypes in comparison to consensus diagnosis [21].
Decision making ability was assessed using the SPACED. This measure is a shortened version of the Assessment of Capacity for Everyday Decision Making (ACED) [22, 23] and evaluates four components of decision making abilities: understanding, appreciation, comparative reasoning, and consequential reasoning. All participants were read the following scenario about a hypothetical problem experienced by a family member: “A family member receives a letter from the electric company saying that she/he has not paid the bill for two years. The electric company says they are going to turn off the power and sue her/him for using the power and not paying during that time. In response, your family member is deciding between either (1) apologizing and paying the full amount on a credit card or (2) leaving the country and avoiding the problem all together.” They are then asked standardized questions to evaluate their understanding and appreciation of the scenario and their ability to engage in comparative and consequential reasoning related to the scenario. These four criteria are scored with a 0 (inadequate), 1 (marginal), or 2 (adequate) generating a possible range of 0 to 8. While the SPACED utilizes a specific scenario, the outcome measure defined by these four core decision making criteria is broadly applicable to a wide range of everyday decisions. The full version of the measure (the ACED) was specifically designed to measure decision making capacity for solving common functional problems encountered by older adults (e.g., medication adherence, management of finances) and has previously been shown to be a reliable and valid measure of decision making [23].
Statistical analyses
Demographic and cognitive battery differences were tested between cognitively normal and MCI groups with two-sample t tests (for continuous variables) and Fisher’s exact tests (for categorical variables). A generalized linear model with a negative binomial distribution was used to examine the relationship between MCI status and decision making (i.e., performance on the SPACED). The overall SPACED score was highly left-skewed, and the most common participant response was an 8 out of 8. In order to model this outcome interpretably and adequately, we reversed the score (by subtracting participant scores from 8) and modeled the number of “points missed” on the SPACED measure as a count outcome in a negative binomial generalized linear model (GLM). The model included our primary explanatory variable, MCI, as a covariate whose coefficient represents our effect of interest. This model also adjusted for age (z-score), education, sex, randomization group, cognitive assessment method, and study site as potential confounding variables. All statistical analyses were performed in R version 4.2.0 [24].
RESULTS
Participant characteristics
The sample consisted of 301 older adults. The mean age was 77.1 years (SD 5.1), 69.4% had a college degree or higher, 51.2% were female, 95.3% were White, and 99.0% were non-Hispanic (Table 1). Participants with MCI had significantly lower screening MoCA scores (p = 0.026). No significant demographic differences were observed between groups. Of note, participants generally performed very well on the SPACED with 77.8% earning a perfect score. When looking by MCI classification, 79.9% of participants classified as cognitively normal earned a perfect score compared to 57.1% of participants classified as MCI.
Participant characteristics
Counts and percentiles or means and standard deviations are shown. CN, cognitively normal; MCI, mild cognitive impairment; CU, Colorado University; IU, Indiana University; UCSD, University of California San Diego; DDA, driving decision aid; BTACT, the Brief Test of Adult Cognition by Telephone; RBANS, the Repeatable Battery for the Assessment of Neuropsychological Status; SPACED, the Short Portable Assessment of Capacity for Everyday Decision making.
MCI classification
A total of 28 participants (9.3%) were categorized as having MCI (N = 16 with BTACT approach; N = 12 with RBANS approach). Notably, there was a significant difference in the percentage of participants categorized as having MCI based on approach (16/237 or 7% with BTACT approach; 12/64 or 18% with RBANS approach). The remaining 269 participants were categorized as cognitively normal. Descriptive statistics for participants’ performances on all cognitive measures are shown in Tables 2 3.
BTACT battery scores (telephone-based assessment)
Means and standard deviations are shown. Two-sample T tests were used to examine differences in continuous variables between CN and MCI groups. Significant differences (p < 0.05) are bolded. CN, cognitively normal; MCI, mild cognitive impairment; BTACT, the Brief Test of Adult Cognition by Telephone.
RBANS battery scores (in-person assessment)
Means and standard deviations are shown. Two-sample T tests were used to examine differences in continuous variables between CN and MCI groups. Significant differences (p < 0.05) are bolded. CN, cognitively normal; MCI, mild cognitive impairment; RBANS, the Repeatable Battery for the Assessment of Neuropsychological Status.
Relationship between MCI and decision making
Results from the negative binomial generalized linear model are presented in Table 4. Adjusting for age, education, sex, randomization group, cognitive assessment method, and study site, MCI status was associated with poorer decision making such that participants categorized as MCI missed an average of 2.17 times more points on the SPACED than those categorized as cognitively normal (mean ratio: 2.17, 95% CI: 1.02, 4.61, p = 0.044).
Model results
Significant differences (p < 0.05) are bolded. SPACED, the Short Portable Assessment of Capacity for Everyday Decision making; CN, cognitively normal; MCI, mild cognitive impairment; DDA, driving decision aid; BTACT, the Brief Test of Adult Cognition by Telephone; RBANS, the Repeatable Battery for the Assessment of Neuropsychological Status. CU, Colorado University; IU, Indiana University; UCSD, University of California San Diego. Asterisks ( *) refer to a joint likelihood ratio test for categorical variables.
DISCUSSION
In the current study of 301 community-dwelling older adults, MCI was associated with poorer performance on the SPACED, a measure designed to evaluate four core components of decision-making abilities: understanding, appreciation, comparative reasoning, and consequential reasoning. This result supports the idea that older adults with MCI exhibit poorer decision making abilities than cognitively normal older adults. Further, because the SPACED is focused on decision making abilities generally rather than any specific category of decisions, this finding suggests that older adults with MCI are likely to experience increased difficulty with a broad range of real-world decisions.
The association between MCI and poorer decision making abilities is largely in line with previous work. MCI has previously been associated with poorer financial and healthcare decision making [6], lower medical decision making capacity [7, 8], lower financial decision making capacity [9], and reduced capacity for consent to treatment or research participation [10–12]. Conversely, differences in decision making abilities between MCI and cognitively normal groups are less consistent using behavioral tasks [9]. The current study is unique in its use of the SPACED to assess decision making abilities. This measure evaluates individuals’ ability to understand a problem, to appreciate why it is a problem, to compare the advantages and disadvantages of two options, and to identify the consequences of each potential choice. Scores therefore reflect practical decision making abilities that extend to a wide array of decisions (e.g., deciding whether to stop driving) as opposed to decision making abilities in discreet circumstances (e.g., financial decision making, research participation).
The finding that MCI is associated with poorer general, everyday decision making abilities as measured by the SPACED has several important implications. First, it suggests that older adults with MCI may benefit from decision aids, particularly for difficult, high stakes decisions. Prior work has shown beneficial effects (e.g., reduced decisional conflict) of decision aids for driving and health related decision making [25–27]. However, very few trials of decision aids have explored whether participants had MCI, highlighting the need for future research in this area [28]. Second, as noted in previous work [12], the current study underscores the importance of conducting accurate evaluations of decisional capacity for individuals with MCI prior to their consent to meaningful decisions (e.g., medical treatments, research participation). Third, the finding highlights the importance of early planning for significant decisions such as creating living wills and designating durable powers of attorney. Decision planning should include discussions of an individual’s goals and values in order to honor those should a time come when their decision making ability is limited [29]. Given prior research showing that even subtle cognitive decline is associated with reduced decision making [1], planning should ideally be done in advance of an MCI diagnosis. Finally, the result of the current study points to the potential utility of shared and supported decision making approaches when working with patients with MCI [30, 31].
Importantly, the finding does not imply that older adults with MCI are incapable of making good decisions independently or that the SPACED should be used to identify individuals with MCI. Rather, it suggests that older adults with MCI exhibit poorer decision-making abilities when compared to cognitively normal older adults and may benefit from additional resources and/or assistance during decision making. However, considering evidence that older adults highly value agency in decision making [32, 33], any support should be offered thoughtfully. For example, in a meta-synthesis of patient preferences for communication with healthcare providers about driving, patients expressed the importance of being in control of their own decisions and receiving concrete, objective evidence of why they should stop driving if instructed to do so [32]. This work highlights the balance between support and respect for autonomy that must be struck when engaging in discussions about decision making with older adults.
There are several limitations to the current study. One notable limitation is the difference in approach used to assess cognitive status across participants. Due to a transition from in-person to telephone-based assessment in response to COVID-19, participants’ cognitive status was measured using either the RBANS or the BTACT and OTMT. While there is evidence to support the method of categorization for each approach, given the significant difference between the percentage of participants categorized as having MCI based on cognitive approach (16/237 or 7% with BTACT approach; 12/64 or 18% with RBANS approach), it is likely that sensitivity and/or specificity of the two measures differed, resulting in inconsistent categorization. For example, the RBANS was likely more sensitive to amnestic MCI, while the BTACT was likely more sensitive to non-amnestic MCI. Given that the BTACT was used to categorize 78.7% of the sample, it is likely that many of the participants in the MCI group have non-amnestic types of MCI. To mitigate issues related to MCI categorization method, we included cognitive assessment method as a covariate when examining the relationship between MCI status and performance on the SPACED. Additionally, given the lack of a premorbid measure of cognitive abilities, it is possible that some participants were categorized as MCI due to lower premorbid cognitive abilities as opposed to a true decline from their baseline abilities. A third limitation is the largely non-Hispanic White, well-educated sample which limits our ability to generalize findings to more diverse populations, or those who live outside the largely urban populations seen at participating centers. Furthermore, the sample consisted of only 28 older adults categorized as MCI. As a result, findings should be interpreted cautiously. Strengths include the use of a decision-making measure not previously associated with MCI and the use of empirically supported MCI diagnostic classifications.
In summary, findings from the study support the idea that MCI is associated with lower performance on a measure of everyday decision making, suggesting that MCI may affect a wide range of decision-making scenarios. However, future work is needed to explore relationships between scores on this measure and functioning in real-world decision-making settings. For example, it remains unclear whether a specific cutoff score on the SPACED may signal clinically relevant impairment in everyday decision making.
Footnotes
ACKNOWLEDGMENTS
The authors have no acknowledgments to report.
FUNDING
This project was supported by NIH/NIA Grant Number R01AG059613 and by NIH/NCATS Colorado CTSA Grant Number UL1 TR002535. Christopher E. Knoepke, PhD, MSW, LCSW is supported by NIH/NHLBI Grant Number K23 HL153892. The contents of this work are the authors’ sole responsibility and do not necessarily represent official funder views or the views of the Department of Veterans Affairs.
CONFLICT OF INTEREST
The authors have no conflict of interest to report.
DATA AVAILABILITY
Requests for data from other researchers and the public will be considered, and data will be made available in accordance with local institution policies, IRB recommendations, local/state/federal laws and regulations, and considerations for publication. Any applicable data sharing will follow HIPAA rules.
