Effects of Healthy Aging and Mild Cognitive Impairment on a Real-Life Decision-Making Task

Abstract

In this study, we investigated the effects of age and of mild cognitive impairment (MCI) on decision making under risk by adopting a task representing real-life health-related situations and involving complex numerical information. Moreover, we assessed the relationship of real-life decision making to other cognitive functions such as number processing, executive functions, language, memory, and attention. For this reason, we compared the performance of 19 healthy, relatively younger adults with that of 18 healthy older adults and the performance of the 18 healthy older adults with that of 17 patients with MCI. Results indicated difficulties in real-life decision making for the healthy older adults compared with the healthy, relatively younger adults. Difficulties of patients with MCI relative to the healthy older adults arose in particular in difficult items requiring processing of frequencies and fractions. Significant effects of age and of MCI in processing frequencies were also evident in a ratio number comparison task. Decision-making performance of healthy participants and of the patient group correlated significantly with number processing. There was a further significant correlation with executive functions for the healthy participants and with reading comprehension for the patients. Our results suggest that healthy older individuals and patients with MCI make less advantageous decisions when the information is complex and high demands are put on executive functions and numerical abilities. Moreover, we show that executive functions and numerical abilities are not only essential in laboratory gambling tasks but also in more realistic and ecological decision situations within the health context.

Keywords

Aging decision making executive functions mild cognitive impairment numeracy

INTRODUCTION

The last decades of life are accompanied by numerous changes that warrant decision making. Older individuals typically have to make decisions related to changes in income, investment of savings, health management, decreased mobility, changes in the family constellation, and often to the loss of near relatives or spouses. The present investigation assesses decision making in realistic health-related situations in healthy participants in advanced age as well as in patients with mild cognitive impairment(MCI).

The term MCI refers to a stadium in between intact cognitive functioning and dementia in older people [1 –3]. Patients with MCI subjectively and objectively experience a decline in cognitive functions greater than that expected for age and education. Most often patients and/or their caregivers complain about diminishing memory functions, but executive functions and other cognitive domains (e.g., attention, visuo-construction, or language) may also be affected. Importantly, patients with MCI are by definition autonomous in daily functioning, but deficits may arise in more complex situations (e.g., [4, 5]). Compared with healthy age-matched controls, patients with MCI have a higher risk for developing dementia, in particular Alzheimer’s disease (AD). One study reported a conversion rate of 29% from MCI to AD within 30 months, but it has also been shown that their minor cognitive deficits can be stable or even disappear over the years [6]. However, it has recently been suggested that a diagnosis of MCI at any time has prognostic value for developing dementia [7]. Several pathologies including AD may be responsible for MCI. Improved diagnostic criteria including biomarkers have been proposed in order to identify MCI due to AD [1].

Previous neuropsychological studies on aging and MCI mainly focused on decision making under risk and decision making under ambiguity using laboratory gambling tasks. In choices under risk, probabilities of different outcomes, possible gains, and possible losses are given or can be calculated, whereas in choices under ambiguity the contingencies of the decision situation are not explicitly stated but have to be learned from feedback [8 –10]. In decisions under risk, performance of older individuals depends on the complexity of the task and on its demands on executive functions. Difficulties have been reported for older individuals in more complex decision under risk tasks [11, 12], while no such effect has been found on a simpler decision task with explicit information about contingencies [13]. Brand and Schiebener [12] showed that the association between age and decision making under risk is moderated by executive functions such as categorization and logical thinking. This means that only those older individuals with insufficient executive functions have problems deciding advantageously under risk. Difficulties in decision making are even more pronounced in patients with MCI [14 –17] or patients with neurodegenerative diseases [18 –21]; for a review see: [22]. As a group, patients with MCI make more sub-optimal choices relative to healthy older controls in risk situations [16, 17]. It has been proposed that a decline in executive functions may affect decision making in healthy aging individuals and in patients with MCI [12 , 24]. Structural and functional changes as well as a decline in the efficacy of various neurotransmitter systems take place in the normal aging brain [25]. Those changes seem to affect particularly the frontal lobes [26 –30]. Structural and functional alterations are more pronounced in patients with MCI than in healthy aging [31], although the nature of the structural and functional alterations depends heavily on the underlying pathology of MCI [32]. Apart from executivefunctions, numerical abilities play an important role in decision making under risk. Basic calculation abilities and more complex numerical processing such as the understanding of percentages or frequencies are found to influence the performance in decision making of healthy individuals [33] and of patients with MCI [16]. Several investigations have reported that certain numerical abilities show a decline with advancing age. Some of these changes may be accounted for by a decline in general cognitive functions such as processing speed or executive functions including flexibility and working memory. Halberda et al. [34] reported a steady decrease from the age of 30 years to the age of 85 years in number acuity, i.e., in the ability to distinguish between the numerosity of two sets of dots. However, this decline has been explained by the interference of non-numerical information that may have a larger effect in older participants [35]. As Cappelletti et al. [35] suggest, numeracy seems to be resilient to aging but to be affected by the decline of inhibitory processes supporting number performance. Peripheral processes may also slow down the encoding and/or answer processes, while simple fact retrieval from memory (e.g., answering 3×3 or 5+2) seems well preserved in older adults [36]. Decreasing attentional and executive functions may have a negative effect in more demanding task situations [36, 37] or in complex mental and written calculation [38] increasing the error rate. Important changes have been documented in strategy selection and strategy execution in advanced age; for a review see [39]. Older adults have a smaller repertoire of strategies, select them less adaptively according to the problems’ characteristics, and execute them less effectively; e.g., [40, 41]. As Uittenhove and Lemaire [39] point out, older adults use strategies in a way that puts minimal demands on their limited cognitive resources, i.e., they maintain fewer strategies active, avoid to switch between strategies, and choose strategies that rely less on working memory and executive functions. Age-related changes have also been reported in ratio processing [42]. Older age, lower education, female gender and lower cognitive functions predicted low ratio processing in a health numeracy task. Executive functions and calculation abilities partly mediated the effect of older age on ratio processing [42].

Difficulties in number processing are more pronounced in patients with MCI than in healthy aging controls. The decrease of attention and executive functions may negatively influence numerical processing in patients with MCI. They have more difficulties than healthy older participants to inhibit irrelevant information in a numerical matching task [36] and to ignore overlearned associations in calculation tasks [37]. Additionally, ratio processing (e.g., converting a percentage into a frequency, indicating which of 3 frequencies represents the highest risk of getting a disease) is more demanding for patients with MCI than for healthy older controls [43].

Decreased numerical abilities have also an impact on the patients’ functioning in activities of everyday life. A recent study by Benavides-Varela et al. [44] observed not only impairments in number comprehension, transcoding, and written operations but also in daily activities involving time estimation and money usage. Association between the competence in handling money and volume of the left mesial frontal cortex, right superior frontal cortex, and right superior temporal cortex were revealed. Overall, correlations were found with brain regions that are not typically associated with numerical representations. As Benavides-Varela et al. [44] suggest, patients with MCI may suffer from a processing deficit involving the frontal lobes bilaterally. Since complex numerical processing requires the coordination of disparate sources and puts high demands on attention and executive functions, computational capacities begin to fail. Other studies observed difficulties of patients with MCI in financial tasks such as financial conceptual knowledge, bank statement management, and bill payment [4 , 45–48]. Griffith et al. [45] reported significant correlations between financial capacity scores and MRI volumes in the group with amnestic MCI. In particular, financial performance was moderately correlated with angular gyrus and precuneus volumes. Angular gyrus volume was found to be a unique predictor of financial abilities after accounting for overall mental status and demographic variables. Finally, the relationship between angular gyrus volume and financial capacity was partially mediated by arithmetic ability. Griffith et al. [45] suggest that early neuropathological modifications in the lateral parietal region would lead to a breakdown of cognitive abilities supporting everyday financial skills.

In sum, several studies reported a decline of executive functions and numerical abilities, which in turn may compromise the older individuals’ capacity to make advantageous decisions under risk. This effect is not only due to the limited performance in doing calculations but may also be attributed to the way numerical information is perceived and elaborated. Following Lipkus and Peters [49], highly numerate people pay more attention to numbers, base their considerations more often on numbers, and rely more on numerical data. One might thus speculate that people with lower numerical competence, caused by MCI or age-related decline, pay less attention to numerical information and do not fully exploit their residual numerical capacities. The present study investigates real-life decision making within the health context and its relationship to the understanding of ratio information. Using a task representing real-life decisions instead of a computerized gambling task should help to extend the knowledge about the decision-making competence of healthy older individuals and especially of patients with MCI. Based on previous findings [42], we assume that healthy older individuals (median age 74 years) experience more difficulties in comparing ratio numbers than healthy, relatively younger individuals (median age 61 years). The comparison of frequency and fraction formats (e.g., 3 out of 100 versus 8 out 1000; 2/3 versus 3/6) may present specific difficulties, as the denominator is often neglected in these formats [50, 51]; see also [52, 53]. Difficulties in comparing ratio numbers should be more pronounced for patients with MCI than for healthy controls [43]. We further hypothesize that relatively younger people perform better in real-life decision making than older individuals [12 , 24], who in turn perform better than patients with MCI [16]. Moreover, we expect that deficits are pronounced in decisions involving complex numerical information (frequencies and fractions). Finally, we assume that performance in real-life decision making is related to both executive functions and ratio processing both in healthy controls and in patients with MCI.

MATERIALS AND METHODS

Participants

For this study, 17 patients (12 female) with MCI and 37 healthy older adults (23 female) were recruited. The patient group had a median age of 79 years (interquartile range, IQR = 73–83), a median education of 10 years (IQR = 09–11), and a median Mini-Mental State Examination (MMSE) score of 27 (IQR = 25–28). For the purpose of this study, healthy controls were split into two groups according to their age. Eighteen controls (age 51 to 69 years) were classified as Age Group I, 19 controls (age 70 to 81 years) as Age Group II. Age group I had a median age of 61 years (IQR = 56–65), median education of 11 years (IQR = 9–13), and a median MMSE score of 30 (IQR = 29–30); Age Group II had a median age of 74 (IQR = 72–79), median education of 10 years (IQR = 8–12), and a median MMSE score of 29 (IQR = 28–30). The three groups did not significantly differ from each other with regard to education, χ²(2, N = 54) = 1.029, p = 0.59, or gender χ²(2, N = 54) = 2.636, p = 0.268. However, there was a significant group effect with regard to the MMSE score, χ²(2, N = 54) = 23.205, p < 0.001. Planned post-hoc contrasts revealed no significant difference between Age Group I and Age Group II, U = 151.00, p = 0.507. As expected, patients with MCI had a significantly lower MMSE score than the healthy older controls from the Age Group II, U = 36.00, p < 0.001. These two groups (MCI versus Age Group II) did not differ from each other in terms of age, U = 119.00, p = 0.186.

Healthy participants were recruited by advertisements from the same local area as patients. All participants underwent a neuropsychological screening (MMSE, CLOX, FAB, digit span forward, and digit span backward; for a description of the tasks, see below), and only those who performed within the normal range were included in the study. None of them had a history of neurological or psychiatric disorders as determined by a screening interview. Patients with MCI were recruited consecutively from the outpatient memory clinic of the Department of Neurology, Medical University of Innsbruck, Austria. They were evaluated prospectively using standard neurological and neuropsychological test procedures. Diagnosis of MCI was based on the core clinical criteria proposed by [1] and [2] (concern about a change in cognition; impairment in one or more cognitive domains 1 ; preservation of independence in functional abilities; no dementia). Based on the neuropsychological assessment and according to the proposed diagnostic scheme by Petersen [2], 6 patients were identified as amnestic MCI single domain (35%), 9 patients as amnestic MCI multi-domain (53%), and the remaining 2 as non-amnestic MCI single domain (12%). These two patients had executive deficits. Patients were autonomous in daily living and reported no impairment in everyday functioning. They underwent neurological and general medical examinations performed by a senior neurologist (TB) as well as laboratory testing to exclude other causes of cognitive decline. The majority of patients were assessed by magnetic resonance imaging (MRI, n = 11) or computed tomography (CT, n = 1). Exclusion criteria were a history of stroke, head trauma, substance abuse, and major medical, psychiatric, or neurological disorders that may compromise cognition. Due to this workup, etiologies such as rapidly progressing dementia, vascular dementia, Levy body dementia, psychiatric disorders, or frontotemporal dementia were largely excluded, as were patients with MCI ‘reverting to normal’. Sixteen out of 17 patients were followed over extended time periods (1 to 4 years); thus, they could be classified as‘stable’ MCI.

Participants did not receive any compensation for their participation in the study. This investigation was approved by the local ethics committee, and written informed consent was obtained from all participants prior to participation.

Comparison of ratio numbers

In this task, participants have to compare 12 pairs of ratio numbers and have to indicate which number represents the higher value. Ratio numbers are presented in different formats. They are either presented as percentages (e.g., 96% versus 88%), fractions (e.g., 1/10 versus 1/25), or frequencies (e.g., 675 out of 1000 versus 8 out of 10). Within the frequency format, the denominator is either given as 10, 100, or 1000. All items are incongruent in the sense that the frequency with the higher operands represents the lower frequency (e.g., 9 out of 10 versus 85 versus 100).

There are 4 items per format, and the items are presented in randomized order. Time limit is set to 30 s per item. For the statistical analysis, we used the total sum of correctly answered items and the subtotal sums of correctly answered items for each format.

Real-life decision-making task

The decision-making task contains 12 short text problems describing situations in a health context (e.g., which physiotherapy to choose, which drug to take ...). Each item consists of 2 alternatives (choice A and choice B), and each alternative contains two pieces of numerical information. One is important for the decision (e.g., success rate of treatment in clinic A or B). The other is rather irrelevant for the decision (e.g., how much patients liked the ambiance of the clinic A or B). The relevant numerical information is always given as percentage, frequency, or fraction. The irrelevant numerical information also contains decimals and integers. Originally, the decision task contained 20 short text problems. For the purpose of validation, these 20 items were presented to 29 healthy young controls (median age = 35.50 years). Only those problems with 100% agreement on the more advantageous alternative were included in the present study. For statistical analysis, the sum of correctly answered questions was used. For a further analysis, we divided the real-life decision-making task into easy and difficult items. In the easy items (n = 9), the relevant numerical information is given in percentages. In the difficult items (n = 3), the relevant numerical information is either given in fractions or in frequencies.

Neuropsychological background tests

All patients performed the CERAD test battery (Consortium to Establish a Registry of Alzheimer’s Disease battery) [54] assessing verbal memory (learning, free recall, and recognition), figural memory (free recall), naming to confrontation (short version of the Boston Naming Test), visuo-construction (copying of geometrical figures), verbal fluency (animals/min, s-words/min), psychomotor speed (Trail Making Test-part A, TMT-A), and cognitive flexibility (TMT-B). This was part of the routine neuropsychological assessment(see Table 1).

Table 1

Medians and interquartile ranges on the CERAD battery for the patient group

	Max.	Patients
	Score	Mdn	IQR
Verbal memory
Learning	30	16.50	13.00–20.00
Free recall	10	4.00	3.00–4.00
Recognition (correct)	20	19.00	18.00–20.00
Figural memory (free recall)	11	6.00	4.00–8.00
Naming to confrontation	15	14.00	12.00–14.00
Visuo-construction (copying of geometrical figures)	11	11.00	10.00–11.00
Semantic verbal fluency (animals/min)		15.00	13.00–20.00
Phonemic verbal fluency (s-words/min)		11.00	9.00–12.00
Psychomotor speed (TMT-A) (s)		59.00	45.00–71.00
Cognitive flexibility (TMT-B/TMT-A)		2.89	2.20–4.35

Mnd, median; IQR, interquartile ranges; CERAD, Consortium to Establish a Registry of Alzheimer’s Disease; TMT, Trail Making Test.

Both patients and controls performed the MMSE of the CERAD battery, the Frontal Assessment Battery (FAB) [55, 56] as well as tests of planning/conceptualization (clock drawing, CLOX1) [57], verbal attention span (digit span forward; Nürnberger Altersinventar, NAI) [58], verbal working memory (digit span backward; NAI) [58], reading comprehension (deciding whether complex statements about five abstract figures are “true” or “false”, also described in [16, 43]), ratio processing (testing comprehension and calculation with ratios, as described in [42, 43]), and mental calculation (addition, subtraction, multiplication, and division problems, as described in [16, 43]). They also responded to a questionnaire on anxiety and depression symptoms (Hospital Anxiety and Depression Scale-German version,HADS-D) [59].

Statistical analysis

All statistical analyses were carried out with IBM SPSS Statistics –Version 24.0 for Windows. Overall, we used non-parametric statistics as the majority of variables did not meet normality assumptions. Tables report median scores, interquartile ranges, Cohen’s d effect-sizes, and significance of Mann-WhitneyU-tests. Group comparisons were made between Age Group I and Age Group II as well as between Age Group II and patients with MCI. A Spearman rank-order correlation analysis was performed between the real-life decision-making task, demographic variables (age, years of education), the ratio number comparison task, and the neuropsychological background tests assessing executive functions (verbal attention span, verbal working memory, planning/conceptualization, FAB total score, cognitive flexibility), ratio processing, mental calculation, and affective symptoms (anxiety and depression). The correlation analysis was performed for the whole control group (Age Group I and II collapsed) and the patient group separately. For all analyses, the alpha level was set at 0.05, and significance was tested2-sided.

RESULTS

Neuropsychological background tests

Age Group II obtained significantly lower scores than Age Group I in the test of verbal attention span, U = 108.00, p = 0.047. There were no significant differences between Age Group I and Age Group II with regard to other neuropsychological background tests. Both groups had comparable anxiety and depression scores (see Table 2).

Table 2

Medians, interquartile ranges, Cohen’s d effect-sizes, and significance of group comparisons in neuropsychological background measures

	Max.	Younger Controls (1)	Older Controls (2)	Patients (3)	1 versus 2	2 versus 3
	score	Mdn (IQR)	Mdn (IQR)	Mdn (IQR)	Cohen’s d	Cohen’s d
Verbal attention span	9	6.50 (6.00–7.00)	6.00 (5.00–7.00)	6.00 (5.00–6.50)	0.633*	0.122
Verbal working memory	8	4.00 (4.00–5.00)	4.00 (4.00–4.00)	4.00 (3.00–4.00)	0.467	0.77*
Conceptualization (CLOX1)	15	13.00 (13.00–14.00)	13.00 (13.00–14.00)	11.00 (10.00–14.00)	0.04	0.589
Frontal Assessment Battery	18	17.00 (17.00–18.00)	16.00 (15.00–18.00)	15.00 (12.50–16.00)	0.623	0.99*
Mental calculation	16	16.00 (15.00–16.00)	15.00 (13.00–16.00)	13.00 (13.00–15.00)	0.544	0.484
Ratio processing	12	9.00 (6.00–11.00)	7.00 (6.00–9.00)	5.00 (5.00–7.00)	0.51	0.542
Reading comprehension	6	6.00 (6.00–6.00)	6.00 (6.00–6.00)	5.00 (5.00–6.00)	0.1	0.911**
Anxiety (HADS-D)	21	5.00 (3.00–8.00)	5.00 (3.00–7.00)	6.00 (3.00–8.00)	0.13	0.159
Depression (HADS-D)	21	4.00 (3.00–5.00)	3.00 (1.00–6.00)	3.00 (2.00–4.00)	0.206	0.132

Group comparisons were performed by means of Mann-Whitney U-test. Mnd, median; IQR, interquartile ranges; HADS-D, Hospital Anxiety and Depression Scale-German version. ^*p < 0.05. ^**p < 0.01.

Compared to Age Group II, patients with MCI obtained significantly lower scores in verbal working memory, U = 93.50, p = 0.027, in the FAB, U = 77.50, p = 0.012, and in reading comprehension, U = 83.00, p = 0.002. Group differences were not significant in the remaining tests, all ps > 0.05. There was also no significant group difference in anxiety and depression scores, both p > 0.05.

Ratio number comparison task

Age Group II showed significantly lower total scores than Age Group I in the ratio number comparison task, U = 99.00, p = 0.025, Cohen’s d = 0.771. Healthy groups significantly differed from each other with regard to the comparison of frequencies, U = 100.00, p = 0.017, Cohen’s d = 0.759, but not with regard to percentages and fractions, both ps > 0.05 (see Fig. 1).

Fig.1

The frequency of correct answers given on the ratio number comparison task is depicted as a function of group (Age Group I, Age Group II, patients with MCI) and number format (percentages, ratios, frequencies). Columns indicate median correct answers, error bars indicate the 25th percentile (lower bar) and 75th percentile (higher bar). ^*p < 0.05.

Patients with MCI obtained significantly lower total scores relative to Age Group II in the ratio number comparison task, U = 97.50, p = 0.041, Cohen’s d = 0.718. Patients significantly differed from Age Group II in the comparison of frequencies, U = 96.00, p = 0.032, Cohen’s d = 0.737, but not in the comparison of percentages and fractions, both p > 0.05 (see Fig. 1).

Real-life decision-making task

Age Group II made overall significantly less advantageous decisions than Age Group I, U = 72.50, p = 0.002, Cohen’s d = 1.13. Contrarily, Age Group II and patients with MCI did not significantly differ from each other with regard to the total number of advantageous decisions, U = 121.00, p = 0.191, Cohen’s d = 0.438.

A post-hoc analysis indicated that Age Group II compared to Age Group I made significantly less advantageous decisions in the easy items, U = 77.00, p = 0.002, Cohen’s d = 1.064. Patients with MCI decided significantly less advantageously than Age Group II in the difficult items, U = 82.00, p = 0.011, Cohen’s d = 0.925 (see Fig. 2). Other contrasts were not significant, allps > 0.05.

Fig.2

Proportion of advantages choices in the real-life decision-making task (median score) is depicted as a function of group (Age Group I, Age Group II, patients with MCI) and item difficulty (easy items, difficult items). Columns indicate median correct answers, error bars indicate the 25th percentile (lower bar) and 75th percentile (higher bar).^*p < 0.05, ^**p < 0.01.

Correlation analysis

As expected, the correlation analysis revealed a significant negative correlation between the performance of healthy controls (n = 37) on the real-life decision-making task and age (r = –0.550, p < 0.001). Real-life decision making significantly and positively correlated with years of education (r = 0.341, p = 0.039), verbal attention span (r = 0.471, p = 0.003), FAB total score (r = 0.506, p = 0.001), mental calculation (r = 0.509, p = 0.001), and total score in the ratio number comparison task (r = 0.461, p = 0.004).

Performance of the patient group on the real-life decision-making task significantly correlated with mental calculation (r = 0.709, p = 0.001), ratio processing (r = 0.525, p = 0.030), and reading comprehension (r = 0.490, p = 0.046). There were also marginally significant correlations with the performance on the ratio number comparison task (r = 0.461, p = 0.063) and executive functions (mental flexibility, r = 0.480, p = 0.051; verbal working memory, r = 0.465, p = 0.069; figural conceptualization, r = 0.444, p = 0.074; word fluency s-words, r = 0.437, p = 0.079). There were no significant correlations with the MMSE score, the remaining tests of the CERAD Battery measuring memory, visuo-construction, or naming to confrontation, or affective symptoms, all ps > 0.1.

DISCUSSION

The main aim of this study was to evaluate the effect of age and of MCI on real-life decision making when ratio numbers are involved. Several cognitive functions including mental calculation, ratio processing, reading comprehension, memory, and executive functions were assessed in order to investigate the relationship between these different cognitive abilities and decision making.

When participants were required to compare ratios and indicate which number represents the higher value, we found that healthy older adults had significantly more difficulties than healthy, relatively younger adults. These difficulties were more pronounced in patients with MCI. The frequency format (e.g., 5 out of 10) posed a special challenge to the healthy older individuals and even more so to the patients with MCI. These problems in ratio processing experienced by healthy older individuals and patients with MCI are in line with results from previous studies [42 , 53]. It should be, however, noted that frequency formats are in general easier to answer than other numerical formats [60]. Frequency items in our investigation were especially difficult as all items were incongruent, in the sense that the expressions with the higher operands represented the lower frequencies (e.g., 9 out of 10 versus 85 out of 100). Thus, this format presented specific difficulties for patients with MCI. Previous investigations have shown an effect of age and of MCI on decision making in computerized gambling tasks [12–17 , 61]. In the current study, healthy controls and patients with MCI were tested on a decision-making task that represents real-life decision situations. In this task, participants had to compare two options, each containing two sentences with numerical information thematically embedded in the health care system. Only one of the two sentences containing numerical information was relevant for the decision (e.g., the rate of successful treatment). We found that healthy older individuals made overall more disadvantageous choices than the relatively younger control group. This finding is in line with results from a previous study by Finucane et al. [62] outlining that older individuals show poorer understanding of information relevant for decision making. Patients with MCI differed from age-matched controls only in difficult items, which means in decisions where the relevant numerical information was given either in frequency or in fraction format. Performance of healthy controls in real-life decision making correlated with demographic variables (age, years of education), executive functions, and number processing (mental calculation, ratio number comparison). Performance of patients with MCI in real-life decision making correlated with number processing (mental calculation, ratio processing) and reading comprehension. Additionally, trend correlations between performance on the decision-making task, ratio number comparison task, and tests of executive functions wererevealed.

Overall, these results confirm previous neuropsychological studies reporting age effects and moderate difficulties of patients with MCI in complex decision tasks [13 , 63]. As demonstrated by the present investigation, these difficulties do not appear only in abstract laboratory task but extend to decisions that have to be mastered in everyday life such as choosing between two possible treatments, between two physiotherapies, or between two insurance companies. One may hypothesize that the present study underestimates the difficulties of patients with MCI in real life as the number of options was limited to two and information was well structured and explicitly given. In real life, the number of options is often far higher. Results of our study converge with previous reports emphasizing the importance of executive functions and ratio processing in decision making under risk. Specifically, it has been suggested that the effect of executive functions on decision making under risk is mediated by ratio processing [33]. As it has been demonstrated in a large sample of healthy participants, good executive functions such as logical reasoning, categorization, flexibility, and working memory positively influenced the processing of ratios and probability knowledge, which in turn led to more advantageous decisions under risk conditions [33]. The present results extend these previous findings suggesting that executive functions and numerical abilities are not only essential in decisions in laboratory gambling tasks but also in more realistic and ecological situations within the health context. Previous studies have shown that patients with MCI may have difficulties in medical decision-making capacity. For example, in a study by Okonkwo et al. [64], patients with MCI performed significantly below the controls’ level in appreciating the consequences of a treatment choice, in providing reasoning for a treatment choice, and in understanding the treatment situation and choices. In follow-up investigations, Okonkwo et al. [64, 65] suggested that special attention should be paid to difficulties in medical decision-making capacity whenever patients with MCI or patients with early AD have to sign informed consent forms or to make medical decisions. Executive functions play an important role in making advantageous decisions, in laboratory gambling tasks and in real-life alike. Thus, patients with MCI who have executive difficulties may experience more pronounced deficits in decision making under risk than patients with MCI of the amnestic type. One might also speculate that, besides age and education, other demographic factors such as gender also influence ratio processing and decision-making abilities. Delazer et al. [42] reported that, in healthy older adults, women had more difficulties than men in a ratio processing task. Thus, lower performance in decision making under risk may be expected in women in advanced age. However, no reliable gender differences were reported in a review on decision making under objective risk conditions [66]. Future studies may address this issue. Furthermore, the effect of different MCI subtypes (e.g., amnestic versus executive MCI) on decision making should be investigated in future research. Health professionals nowadays increasingly try to involve patients in treatment decisions and into their health management, and to respect their autonomous and informed decisions. However, older people often prefer to take a passive role in the decision process [62 , 68]. Results of the present study suggest that healthy older individuals and patients with MCI make less advantageous decisions when the information is complex and high demands are put on executive functions and numerical abilities. These difficulties possibly contribute to old people’s tendency to refrain from actively participating in treatment decision making [67]. Creating a more favorable decision situation (e.g., by using simple numerical information, simple text, or pictograms) may enable persons with slight cognitive deficits to actively manage their health decisions and to participate in eventual treatment processes as cooperative and informedpartner.

Footnotes

Impairments were defined as scores below or equal to –1.5 SD from age-, education-, and gender-scaled norms. Each single participant in the patient group had at least one deficit reflected by a score below –1.5 SD.

ACKNOWLEDGMENTS

We would like to thank all participants for their contribution to the study as well as Heidi Loos for her help in the data collection with healthy participants. This work is part of the first author’s doctoral thesis and was supported by a fellowship granted by the “Vizerektorat für Forschung” of the Leopold-Franzens-University Innsbruck, 2014/3/PSY-15. L.Z. receives research support from TWF-2010-1-993 and MUI-Start 2014-05-001.

Authors’ disclosures available online ().

References

Albert

, DeKosky

, Dickson

, Dubois

, Feldman

, Fox

, Gamst

, Holtzman

, Jagust

, Petersen

, Snyder

, Carrillo

, Thies

, Phelps

(2011) The diagnosis of mild cognitive impairment due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 270–279.

Petersen

(2004) Mild cognitive impairment as a diagnostic entity. J Intern Med 256, 183–194.

Winblad

, Palmer

, Kivipelto

, Jelic

, Fratiglioni

, Wahlund

, Nordberg

, Backman

, Albert

, Almkvist

, Arai

, Basun

, Blennow

, de

, DeCarli

, Erkinjuntti

, Giacobini

, Graff

, Hardy

, Jack

, Jorm

, Ritchie

, van

, Visser

, Petersen

(2004) Mild cognitive impairment–beyond controversies, towards a consensus: Report of the International Working Group on Mild Cognitive Impairment. J Intern Med 256, 240–246.

Griffith

, Belue

, Sicola

, Krzywanski

, Zamrini

, Harrell

, Marson

(2003) Impaired financial abilities in mild cognitive impairment: A direct assessment approach. Neurology 60, 449–457.

Okonkwo

, Wadley

, Griffith

, Belue

, Lanza

, Zamrini

, Harrell

, Brockington

, Clark

, Raman

, Marson

(2008) Awareness of deficits in financial abilities in patients with mild cognitive impairment: Going beyond self-informant discrepancy. Am J Geriatr Psychiatry 16, 650–659.

Fischer

, Jungwirth

, Zehetmayer

, Weissgram

, Hoenigschnabl

, Gelpi

, Krampla

, Tragl

(2007) Conversion from subtypes of mild cognitive impairment to Alzheimer dementia. Neurology 68, 288–291.

Roberts

, Knopman

, Mielke

, Cha

, Pankratz

, Christianson

, Geda

, Boeve

, Ivnik

, Tangalos

, Rocca

, Petersen

(2014) Higher risk of progression to dementia in mild cognitive impairment cases who revert to normal. Neurology 82, 317–325.

Brand

, Labudda

, Markowitsch

(2006) Neuropsychological correlates of decision-making in ambiguous and risky situations. Neural Netw 19, 1266–1276.

Hsu

, Bhatt

, Adolphs

, Tranel

, Camerer

(2005) Neural systems responding to degrees of uncertainty in human decision-making. Science 310, 1680–1683.

10.

Tobler

, Weber

(2014) Valuation for risky and uncertain choices. In Neuroeconomics: Decision Making and the Brain, Glimcher

, Fehr

, Eds. Elsevier, Amsterdam, pp. 149–172.

11.

Brand

, Markowitsch

(2010) Aging and decision-making: A neurocognitive perspective. Gerontology 56, 319–324.

12.

Brand

, Schiebener

(2013) Interactions of age and cognitive functions in predicting decision making under risky conditions over the life span. J Clin Exp Neuropsychol 35, 9–23.

13.

Zamarian

, Sinz

, Bonatti

, Gamboz

, Delazer

(2008) Normal aging affects decisions under ambiguity, but not decisions under risk. Neuropsychology 22, 645–657.

14.

Bayard

, Jacus

, Raffard

, Gely-Nargeot

(2014) Apathy and emotion-based decision-making in amnesic mild cognitive impairment and Alzheimer’s disease. Behav Neurol 2014, 231469.

15.

Bayard

, Jacus

, Raffard

, Gely-Nargeot

(2015) Conscious knowledge and decision making under ambiguity in mild cognitive impairment and Alzheimer disease. Alzheimer Dis Assoc Disord 29, 357–359.

16.

Pertl

M-T

, Benke

, Zamarian

, Delazer

(2015) Decision making and ratio processing in patients with mild cognitive impairment. J Alzheimers Dis 48, 765–779.

17.

Zamarian

, Weiss

, Delazer

(2011) The impact of mild cognitive impairment on decision making in two gambling tasks. J Gerontol B Psychol Sci Soc Sci 66, 23–31.

18.

Brand

, Labudda

, Kalbe

, Hilker

, Emmans

, Fuchs

, Kessler

, Markowitsch

(2004) Decision-making impairments in patients with Parkinson’s disease. Behav Neurol 15, 77–85.

19.

Delazer

, Sinz

, Zamarian

, Benke

(2007) Decision-making with explicit and stable rules in mild Alzheimer’s disease. Neuropsychologia 45, 1632–1641.

20.

Delazer

, Sinz

, Zamarian

, Stockner

, Seppi

, Wenning

, Benke

, Poewe

(2009) Decision making under risk and under ambiguity in Parkinson’s disease. Neuropsychologia 47, 1901–1908.

21.

Stout

, Rodawalt

, Siemers

(2001) Risky decision making in Huntington’s disease. J Int Neuropsychol Soc 7, 92–101.

22.

Gleichgerrcht

, Ibanez

, Roca

, Torralva

, Manes

(2010) Decision-making cognition in neurodegenerative diseases. Nat Rev Neurol 6, 611–623.

23.

Denburg

, Tranel

, Bechara

(2005) The ability to decide advantageously declines prematurely in some normal older persons. Neuropsychologia 43, 1099–1106.

24.

Denburg

, Cole

, Hernandez

, Yamada

, Tranel

, Bechara

, Wallace

(2007) The orbitofrontal cortex, real-world decision making, and normal aging. Ann N Y Acad Sci 1121, 480–498.

25.

, Rieckmann

(2014) Neuromodulation and aging: Implications of aging neuronal gain control on cognition. Curr Opin Neurobiol 29, 148–158.

26.

Hedden

, Gabrieli

(2004) Insights into the ageing mind: A view from cognitive neuroscience. Nat Rev Neurosci 5, 87–96.

27.

Resnick

, Pham

, Kraut

, Zonderman

, Davatzikos

(2003) Longitudinal magnetic resonance imaging studies of older adults: A shrinking brain. J Neurosci 23, 3295–3301.

28.

West

(1996) An application of prefrontal cortex function theory to cognitive aging. Psychol Bull 120, 272–292.

29.

West

(2000) In defense of the frontal lobe hypothesis of cognitive aging. J Int Neuropsychol Soc 6, 727–729.

30.

West

, Murphy

, Armilio

, Craik

, Stuss

(2002) Lapses of intention and performance variability reveal age-related increases in fluctuations of executive control. Brain Cogn 49, 402–419.

31.

Driscoll

, Davatzikos

, An

, Wu

, Shen

, Kraut

, Resnick

(2009) Longitudinal pattern of regional brain volume change differentiates normal aging from MCI. Neurology 72, 1906–1913.

32.

Petersen

, Caracciolo

, Brayne

, Gauthier

, Jelic

, Fratiglioni

(2014) Mild cognitive impairment: A concept in evolution. J Intern Med 275, 214–228.

33.

Brand

, Schiebener

, Pertl

M-T

, Delazer

(2014) Know the risk, take the win: How executive functions androbability processing influence advantageous decision making under risk conditions. J Clin Exp Neuropsychol 36, 914–929.

34.

Halberda

, Ly

, Wilmer

, Naiman

, Germine

(2012) Number sense across the lifespan as revealed by a massive Internet-based sample. Proc Natl Acad Sci U S A 109, 11116–11120.

35.

Cappelletti

, Didino

, Stoianov

, Zorzi

(2014) Number skills are maintained in healthy ageing. Cogn Psychol 69, 25–45.

36.

Zamarian

, Stadelmann

, Nurk

, Gamboz

, Marksteiner

, Delazer

(2007) Effects of age and mild cognitive impairment on direct and indirect access to arithmetic knowledge. Neuropsychologia 45, 1511–1521.

37.

Zamarian

, Semenza

, Domahs

, Benke

, Delazer

(2007) Alzheimer’s disease and mild cognitive impairment: Effects ofshifting and interference in simple arithmetic. J Neurol Sci 263, 79–88.

38.

Delazer

, Girelli

, Granà

, Domahs

(2003) Numberprocessing and calculation–normative data from healthy adults. Clin Neuropsychol 17, 331–350.

39.

Uittenhove

, Lemaire

(2015) Numerical cognition during cognitive aging. In The Oxford Handbook of Numerical Cognition, Cohen Kadosh

, Dowker

, Eds. Oxford University Press, Oxford, pp. 345–364.

40.

Hodzik

, Lemaire

(2011) Inhibition and shifting capacitiesmediate adults’ age-related differences in strategy selection andrepertoire. Acta Psychol (Amst) 137, 335–344.

41.

Lemaire

, Arnaud

(2008) Young and older adults’ strategies in complex arithmetic. Am J Psychol 121, 1–16.

42.

Delazer

, Kemmler

, Benke

(2013) Health numeracy and cognitive decline in advanced age. Neuropsychol Dev Cogn B Aging Neuropsychol Cogn 20, 639–659.

43.

Pertl

M-T

, Benke

, Zamarian

, Martini

, Bodner

, Karner

, Delazer

(2014) Do patients with mild cognitive impairment understand numerical health information? J Alzheimers Dis 40, 531–540.

44.

Benavides-Varela

, Burgio

, Meneghello

, De

, Arcara

, Rigon

, Pilosio

, Butterworth

, Venneri

, Semenza

(2015) Anatomical substrates and neurocognitive predictors of daily numerical abilities in mild cognitive impairment. Cortex 71, 58–67.

45.

Griffith

, Stewart

, Stoeckel

, Okonkwo

, den Hollander

, Martin

, Belue

, Copeland

, Harrell

, Brockington

, Clark

, Marson

(2010) Magnetic resonance imaging volume ofthe angular gyri predicts financial skill deficits in people withamnestic mild cognitive impairment. J Am Geriatr Soc 58, 265–274.

46.

Marson

, Martin

, Wadley

, Griffith

, Snyder

, Goode

, Kinney

, Nicholas

, Steele

, Anderson

, Zamrini

, Raman

, Bartolucci

, Harrell

(2009) Clinical interview assessment of financial capacity in older adults with mild cognitive impairment and Alzheimer’s disease. J Am Geriatr Soc 57, 806–814.

47.

Okonkwo

, Griffith

, Vance

, Marson

, Ball

, Wadley

(2009) Awareness of functional difficulties in mild cognitive impairment: A multidomain assessment approach. J Am Geriatr Soc 57, 978–984.

48.

Sherod

, Griffith

, Copeland

, Belue

, Krzywanski

, Zamrini

, Harrell

, Clark

, Brockington

, Powers

, Marson

(2009) Neurocognitive predictors of financial capacity across the dementia spectrum: Normal aging, mild cognitive impairment, and Alzheimer’s disease. J Int Neuropsychol Soc 15, 258–267.

49.

Lipkus

, Peters

(2009) Understanding the role of numeracy inhealth: Proposed theoretical framework and practical insights. Health Educ Behav 36, 1065–1081.

50.

Garcia-Retamero

, Okan

, Cokely

(2012) Using visual aids to improve communication of risks about health: A review. Scientific World Journal 2012, 562637.

51.

Reyna

, Brainerd

(2008) Numeracy, ratio bias, and denominator neglect in judgments of risk and probability. Learn Individ Differ 18, 89–107.

52.

Iuculano

, Butterworth

(2011) Understanding the real value of fractions and decimals. Q J Exp Psychol (Hove) 64, 2088–2098.

53.

Millroth

, Juslin

(2015) Prospect evaluation as a function of numeracy and probability denominator. Cognition 138, 1–9.

54.

Berres

, Monsch

, Bernasconi

, Thalmann

, Stahelin

(2000) Normal ranges of neuropsychological tests for the diagnosis of Alzheimer’s disease. Stud Health Technol Inform 77, 195–199.

55.

Dubois

, Slachevsky

, Litvan

, Pillon

(2000) The FAB: A Frontal Assessment Battery at bedside. Neurology 55, 1621–1626.

56.

Benke

, Karner

, Delazer

(2013) FAB-D: German version of the Frontal Assessment Battery. J Neurol 260, 2066–2072.

57.

Royall

, Cordes

, Polk

(1998) CLOX: An executive clock drawing task. J Neurol Neurosurg Psychiatry 64, 588–594.

58.

Oswald

, Fleischmann

(1997) Nürnberger-Alters-Inventar, Hogrefe.

59.

Herrmann

, Buss

, Lingen

, Kreuzer

(1994) [The screening for anxiety and depression in routine medical care]. Dtsch Med Wochenschr 119, 1283–1286.

60.

Gigerenzer

(1996) The psychology of good judgment: Frequency formats and simple algorithms. Med Decis Making 16, 273–280.

61.

Fein

, McGillivray

, Finn

(2007) Older adults make less advantageous decisions than younger adults: Cognitive and psychological correlates. J Int Neuropsychol Soc 13, 480–489.

62.

Finucane

, Slovic

, Hibbard

, Peters

, Mertz

, Macgregor

(2002) Aging and decision-making competence: An analysis of comprehension and consistency skills in older versus younger adults considering health-plan options. J Behav Decis Mak 15, 141–164.

63.

Denburg

, Cole

, Hernandez

, Yamada

, Tranel

, Bechara

, Wallace

(2007) The orbitofrontal cortex, real-world decision making, and normal aging. Ann N Y Acad Sci 1121, 480–498.

64.

Okonkwo

, Griffith

, Belue

, Lanza

, Zamrini

, Harrell

, Brockington

, Clark

, Raman

, Marson

(2007) Medical decision-making capacity in patients with mild cognitive impairment. Neurology 69, 1528–1535.

65.

Okonkwo

, Griffith

, Copeland

, Belue

, Lanza

, Zamrini

, Harrell

, Brockington

, Clark

, Raman

, Marson

(2008) Medical decision-making capacity in mild cognitive impairment: A 3-year longitudinal study. Neurology 71, 1474–1480.

66.

Schiebener

, Brand

(2015) Decision making under objective risk conditions - a review of cognitive and emotional correlates, strategies, feedback processing, and external influences. Neuropsychol Rev 25, 171–198.

67.

Deber

, Kraetschmer

, Urowitz

, Sharpe

(2007) Do people want to be autonomous patients? Preferred roles in treatment decision-making in several patient populations. Health Expect 10, 248–258.

68.

Hibbard

, Slovic

, Peters

, Finucane

, Tusler

(2001) Is the informed-choice policy approach appropriate for Medicare beneficiaries? Health Aff (Millwood) 20, 199–203.