The domain of FDM was analyzed as part of a larger scoping review. The primary aim was to identify existing decision-making measures applied to adults in midlife and later. Further protocol details can be found in the protocol of the parent multiple-domain scoping review (Ho et al., 2024). This study is exempt and classified as non-human subjects research at Northwestern University (STU00220334).

Upon establishing the research questions (Ho et al., 2024), we developed a search strategy and inclusion criteria consistent across all domains of the multi-domain scoping review. The eligibility criteria encompassed cohort studies, case–control studies, and randomized control trials that evaluated DM in adults aged 45 and above (see Supplementary Table S1 for details). Previous scoping reviews on aging and DM ability have focused on later life, typically on individuals aged 50 or 60 (e.g., Mah et al., 2021; Raj et al., 2019; Usher and Stapleton, 2022; Walsh et al., 2017). To identify tools sensitive to early changes in cognitive aging, we included articles with study populations starting from midlife (i.e., ages 45 and above). Studies on shared decision-making, decision aids, and perceptual decision-making tasks (e.g., the random dot motion task; von Lautz et al., 2019) were excluded from consideration.

We conducted a systematic search across multiple databases, including EMBASE (Elsevier), PsycINFO, PubMed, MEDLINE, PsychARTICLES, and Web of Science, to identify relevant studies published between January 1, 2018 and November 06, 2023. This timeframe was meant to capture measures currently used in clinical and research settings and those that could be easily adapted to a digital format. The scoping review used several search terms to capture three broad categories: decision-making (e.g., “decisional impairment”), financial (e.g., “financial management”), and measurement (e.g., “assessment,” “tool”). See Supplementary Table S2 for search terms specific to FDM.

The scoping review proceeded through three phases: (1) title and abstract screening, (2) full-text review, and (3) full-text extraction and synthesis.

The initial database search yielded 32,235 articles based on the search criteria (Figure 1). After 15,957 duplicates were removed, the remaining 16,278 articles were reviewed in Covidence (Veritas Health Innovation, 2024). The remaining articles were first screened by title and abstract, resulting in the exclusion of 14,622 articles. Following full-text screening, 869 additional articles were excluded. The remaining 787 articles advanced to the extraction phase, during which 82 more articles were excluded. Reasons for exclusion at this stage included articles that tested paradigms tangential to FDM such as reinforcement learning paradigms (Biernacki et al., 2020; Brown et al., 2020), which were determined to be unsuitable for a clinical screening context or relied only on clinician evaluation with no mention of the standardized qualitative or quantitative measure used (e.g., Gan et al., 2023). This resulted in a final set of 705 articles in the overall multi-domain scoping review. Of these 705 articles, 154 focused on FDM in adults 45 and older. Overall, 42.2% of these 154 articles focused solely on FDM, while 57.8% covered multiple domains. The full list of these 154 articles is provided in Supplementary Table S4. Finally, the PRISMA flow chart for all these stages of the current scoping review is presented in Figure 1.

Among the 154 articles focusing on the FDM domain, we identified 96 unique FDM measures. A large portion (67.7%) of these 96 measures were used only once across the sample, either because they were specifically designed for the study or represented a variation of a paradigm. On the other hand, some measures were used in more than one study, resulting in the total frequency of 227 for these 96 FDM measures. For example, the Iowa Gambling Task (IGT; Bechara et al., 1994) was used in 22 studies. Frequency rate for each FDM measure is provided in Supplementary Table S5.

The five most commonly used measures to assess FDM were the Iowa Gambling Task (IGT; Bechara et al., 1994), Legal Capacity for Property Law Transactions Assessment Scale (LCPLTAS; Giannouli et al., 2018), Decision-making Competence Assessment Tool (DMCAT; Finucane and Gullion, 2010), variations of the temporal discounting task (Benzion et al., 1989; Green et al., 1994; McClure et al., 2007; McHugh and Wood, 2008; Sellitto et al., 2010), and the Financial Capacity Inventory (FCI; FCI-SF; Marson et al., 2000). These five FDM measures are briefly described below. In addition, the list of the 15 most frequent FDM measures is presented in Table 1 (see Supplementary Table S5 for the entire list of the 96 FDM measures).

Among the DM measures identified in the current scoping review, the Lichtenberg Financial Decision Rating Scale (LFDRS; Lichtenberg et al., 2015) appeared to be the most comprehensive one, addressing DM ability alongside financial knowledge and instrumental activities of daily living (IADLs) associated with finances. It is designed in a multiple-choice format, covering key concepts essential to DM ability, such as financial situational awareness, psychological vulnerability, financial exploitation, undue influence, and current financial transactions. The LFDRS was used in 1.8% of the studies included in this scoping review. In every case, the measure was administered by an examiner. Typically, this measure was administered in-person (75%) though one article did not provide mode of administration (25%). The measure was not used in any specific clinical group, however, in all instances, it was administered to adults aged 60 and above.

We extracted information for all unique FDM measures identified on their (1) modality of administration (i.e., remote vs. in-person), (2) self-administration and supervision, and (3) the technology (e.g., computer, tablet) used, (4) duration of administration, and (5) psychometric evidence (i.e., reliability and validity).

Modality of administration. Most of the measures (74%) were conducted in-person. Remote and hybrid administration of the measures was less common; remote administration accounted for 15.6% of the total measures and hybrid administration accounted for 5.2% of the total measures. Modality of administration was not specified for the remaining 5.2% of the measures.

Examiner or self-administration. Approximately a quarter (25.1%) were self-administered under supervision, and 26.4% were reported as self-administered with no examiner supervision. More than a third (38.3%) of the measures were administered by an examiner or clinician. For the remaining 10.1% of measures, no information was reported on self-administration and supervision.

Technology format. Nearly half of the measures (47.6%) required some form of technology, most commonly a computer. Of 227 applications of 96 DM measures reviewed in this article, 8.8% were administered via computer and remotely. The remaining studies either used paper/pen (5.3%) or did not report any use of technology (47.1%).

Duration of administration. Most articles did not report on the measures’ duration to completion (91.2%) and psychometric properties of internal consistency (87.6%), test–retest reliability (98.5%), and inter-rater reliability (96.4%).

Studies included in the present scoping review were examined regarding their sample characteristics (e.g., age range, diagnosis, and language). Among all reviewed studies, 93.5% explicitly reported that they involved participants aged 45 and older, while the remaining 6.5% could have been inferred for older populations based on their clinical diagnosis (e.g., mild cognitive impairment (MCI); dementia of the AD type (predominantly amnestic); posterior cortical atrophy, a syndrome in which visuospatial decline precedes other types of cognitive impairment) which typically affects populations over the age of 45. Across all studies, 70.8% included participants over the age of 65. With respect to the inclusion of clinical samples, 40.9% of 154 studies did not include a clinical group, 40.9% included both clinical samples and healthy controls, and 18.2% included only clinical samples (Figure 2). Of the 154 studies included in this scoping review, 13% included multiple clinical groups (e.g., participants with schizophrenia and Huntington’s disease). The most common clinical populations were patients with a diagnosis of mild cognitive impairment (MCI) due to Alzheimer’s Disease (AD) as well as other forms of dementia (e.g., frontotemporal, vascular) in 24% of the studies. Additionally, 21.4% of the studies included samples with psychiatric disorders (e.g., bipolar disorder, schizophrenia, major depressive disorder, obsessive-compulsive disorder, and various behavioral addictions and substance use disorders). Movement disorders, including Parkinson’s disease (PD), Huntington’s disease (HD), and Amyotrophic lateral sclerosis (ALS) were mentioned in 7.8% of the articles. Neurologic conditions (e.g., stroke, multiple sclerosis, brain lesions) were mentioned in 7.1% and various health conditions (e.g., HIV, cancer, frailty) were mentioned in 1.9% of articles. Finally, results regarding the language of administration in the study demonstrated that approximately half of the studies were conducted in English (53.9%), while the rest were conducted in other languages, such as Greek (13%), Italian (8.4%), German (6.5%), Dutch (2.6%), and Japanese (3.2%).

Many older adults with cognitive impairment—as high as 7.4 million older adults in the US, according to one study (Li et al., 2022)—are managing their household finances, increasing the likelihood of adverse financial consequences or mismanagement. Nicholas et al. (2021) showed, in a large cohort study of more than 80 thousand Medicare beneficiaries living alone, that those with ADRD were more likely to miss bill payments up to 6 years prior to a formal clinician diagnosis, and Lichtenberg et al. (2024) show that earlier memory loss predicts excess spending. These recent findings suggest there are potentially highly detectable indicators, even while an individual presents as cognitively normal. Furthermore, those unaware of their cognitive decline are more likely to suffer financial loss (Mazzonna and Peracchi, 2024), highlighting the gains that can be achieved with earlier detection.

This scoping review identified numerous measures for detecting and characterizing aspects of financial decision-making decline. Future research should aim to build upon these existing tools so they can be implemented as individuals transition from middle to late adulthood, potentially making early identification of FDM deficits a possibility. Earlier detection of financial vulnerability may allow for earlier entry points for individuals to be targeted for a variety of interventions, such as those related directly to potential dementia diagnosis and treatment, as well as functional collateral consequences. For example, evidence of early potential financial mismanagement may prompt clinicians or family members to seek a more thorough neuropsychological evaluation for the individual. Recently tested interventions, such as the randomized clinical trial Plan Your Lifespan, showed that a knowledge-based, interactive intervention increased understanding, accessing, and planning for health decisions for aging in place in cognitively healthy adults aged 65 and above (Lindquist et al., 2017). Early planning, for example, of intra-family-member transfers of FDM responsibilities, can help reduce burden on the family and potentially on public resources. Early detection of signs of financial vulnerability with standardized screening measures may serve as a timely prompt for a variety of decisions related to household finance planning, ultimately preserving financial independence and autonomy for as long as possible.

There is also a growing call to better understand, characterize, and sensitively measure the behavioral and social risk factors that predict ADRD manifestation (Harrell et al., 2024; Martínez-Nicolás et al., 2021). Advances in health measurement in this area include, for example, the development and validation of smartphone-based ecological momentary testing paradigms such as the NIA-funded Mobile Toolbox project (Gershon et al., 2022; Landavazo et al., 2023; Rentz et al., 2024), passive digital signatures gleaned from sources like electronic data warehouses and medical records (e.g., Boustani et al., 2020) or remote home or individual sensing technologies such as accelerometry through real-world navigation tasks.

There may also be additional diagnostic value from information procured from caregivers, whether they be from standardized questionnaires (Olde Rikkert et al., 2011) or new technologies and assessment paradigms that have a growing evidence-base in dementia patients, and could easily be applied to caregivers, such as in-home sensors. For example, a study looking at remote sensors in homes of dementia patients (Gaugler et al., 2019) suggests that movement patterns in the home may be able to detect early impairment.

Finally, these measurement development and validation efforts should be informed by specific individual circumstances, such as extant household assets (Lee et al., 2024), norms surrounding intra-household exchange and sharing of financial resources (Bertocchi et al., 2014), and other social determinants of health and cultural norms and values regarding how financial decisions are made. Ultimately, personalized approaches using a combination of theoretically and data-guided indices may enhance measurement precision, greatly aiding in mitigating the severe economic burdens of age-related neurodegenerative disease.