An account of personal autonomy for people living with dementia

Abstract

Background and Objectives

People living with dementia see autonomy as central to their well-being, and loss of autonomy is one of the things people diagnosed with dementia fear the most. Effective support of autonomy requires us to understand carefully what autonomy is and to structure care plans and health policy in accordance with that understanding. Many recent social scientific studies of autonomy in people with dementia do not carefully operationalize the term “autonomy.” This is problematic because autonomy is a highly ambiguous term that points to a complex reality. We distill from the relevant philosophical and empirical literatures a conceptually clear and empirically informed account that is relevant to the experience of people living with dementia.

Research Design and Methods

We present a general account of the concept of personal autonomy, drawing on existing philosophical literature. We then test the relevance of this account via a scoping review of empirical research reporting on the experience of personal autonomy in persons living with dementia.

Results

With the assumption that adequate relational supports are in place, all aspects of our philosophically informed account of personal autonomy (decisional autonomy, authenticity, and executional autonomy) are compatible with the experience of persons living with dementia.

Discussion and Implications

With adequate relational support, personal autonomy remains an achievable goal even in contexts of moderate to severe dementia. A conception of personal autonomy that is both theoretically and empirically informed can help guide efforts to study and support personal autonomy in persons living with dementia.

People living with dementia see autonomy as central to their well-being and greatly fear losing autonomy (Macquarrie, 2005; O’Rourke et al., 2015). Support for personal autonomy is central to the ideal of person-centered care (Fazio et al., 2018). But supporting the autonomy of persons living with dementia requires understanding what autonomy is and structuring care plans and health policy accordingly. Many recent studies of autonomy in people with dementia do not carefully operationalize the term “autonomy.” This is problematic because autonomy is an ambiguous term that points to a complex reality (Collopy, 1988). This article addresses the problem by presenting a theoretical account of personal autonomy, drawing on existing philosophical literature, and by testing this account against empirical research reporting on the experience of persons living with dementia. Dementia involves the progressive deterioration of cognitive functions needed to exercise personal autonomy, such as memory recall and formation, thinking, judgment, comprehension, calculation, orientation, learning, and language (WHO, 1992). It can seem futile, therefore, to attempt to support the autonomy of persons with dementia. We resist this hopeless approach, which risks justifying a substandard provision of care, suggesting instead that personal autonomy is achievable even in contexts of moderate to severe dementia. But careful thought is required to see how persons living with dementia can continue to exercise autonomy despite their changing situation. This article distills from the philosophical and social-scientific literatures a conceptually clear and empirically informed account that will help guide future efforts to study and support personal autonomy in persons living with dementia.

Part 1: a concept-map of personal autonomy

Philosophers have devoted many thousands of pages to analyzing the concept of personal autonomy. We begin by articulating, in simple terms, what these theorists mean by “personal autonomy.” In the interest of making available something useful to empirical researchers and health practitioners, we gloss over several areas of debate, focusing on common ground among theorists of personal autonomy. For a description of our method for surveying this literature, see online supplementary material Appendix C. To begin, it is important to distinguish personal autonomy as a capacity from autonomy as a right (e.g., the right to refuse treatment) and from autonomy as a moral ideal. For instance, Immanuel Kant’s seminal account of autonomy as rational self-legislation (giving oneself the moral law through reason) is arguably a normative and moral conception of autonomy rather than a descriptive theory of personal autonomy in the contemporary sense. Without denying the importance of other uses of “autonomy,” our focus is on autonomy as a capacity. As our review shows, the philosophical literature standardly views personal autonomy as the capacity or ability to govern, direct, or determine one’s own life. Autonomy in the sense of the capacity for self-determination is particularly relevant to the goal of providing high-quality person-centered care for persons living with dementia. The remainder of this part elaborates on the concept of personal autonomy expressed in the following preliminary definition:

Personal Autonomy: A person is autonomous if and only if they are able to do what they (truly) want to do.

The above definition summarizes the conception of personal autonomy we derive from the relevant philosophical literature. This conception centers personal autonomy on action; it is the ability to act intentionally and to bring about certain outcomes by acting. But not all intentional actions are autonomous. Personal autonomy involves intentional action that accords with the way the person truly wishes to act. We act intentionally, but not autonomously, when internal or external pressures lead us to act in ways we would not ideally choose. Further, given our embodied and socially embedded nature, success in acting in the way we want to act is always partially determined by factors external to ourselves. Personal autonomy as an ability to act, therefore, includes some measure of influence over aspects of one’s environment.

As we will show, the theoretical literature suggests a division of personal autonomy into an inward aspect and an outward aspect, each aspect consisting of a nexus of interconnected abilities. Mackenzie (2021) presents an account similar to our own, distinguishing “internalist” and “externalist” aspects of autonomy and arguing that these are best understood as “causally interconnected dimensions or axes of autonomy” (p. 381). Ideal autonomy would require all of the capacities necessary for personal autonomy—physical, cognitive, emotional, socio-relational, economic, etc.—operating at full power without hindrance. It is unlikely, however, that any human being ever attains perfect autonomy, since we typically face limitations in many or all these areas. Thus, the abilities that together constitute personal autonomy should be envisioned as sliding scales; each person’s capacities function at different levels on the various measures, and our levels can change from season to season, day to day, and even hour to hour. Personal autonomy can be compared to the overall sound output that results from the various inputs and settings on a soundboard. As the sound coming from the speakers is constituted by numerous inputs for vocals and instruments, each with their own levels and settings, so our overall ability to act autonomously is affected by the state and functional level of numerous sub-capacities. To require the perfect operation of all these sub-capacities as a condition for possession of personal autonomy would imply the practically absurd conclusion that none of us is ever autonomous. For this reason, it is more helpful to view the possession of personal autonomy as a matter of degree rather than an either-or, since it depends on a wide variety of sub-capacities, each of which may be operating more or less well at any given moment. Consequently, each of us has the ability to do what we truly want to a greater or lesser extent. As noted, the capacities involved in personal autonomy can be grouped into two main categories: those constituting inward autonomy and those constituting outward autonomy. In what follows, we unpack this preliminary conception of personal autonomy to display the complex set of capacities that constitute the ability to do what one truly wants to do.

Inward autonomy

Theorists of personal autonomy divide into those who focus on the internal psychological capacities of the agent (internalists) and those who emphasize factors external to the individual, such as socio-relational status (externalists). Following Mackenzie (2021), we believe that both approaches are important and that each highlights a different aspect of personal autonomy. Accordingly, we distinguish inward from outward autonomy. Inward autonomy refers to a set of capacities that enables us to form decisions, set goals, and make choices. Among these internal capacities, we can distinguish between those involved in making decisions (decisional autonomy) and those that enable us to recognize decisions as “our own” (to be discussed in the next section on authenticity). Decisional autonomy consists of the ability to decide for oneself what one wishes to do. Many of the mundane decisions we make do not require much in the way of conscious deliberation, but more difficult decisions, or those with higher stakes, typically require numerous sensory and cognitive functions, not to mention the role of emotions and values. Making a decision can involve comparing and evaluating options, weighing and testing evidence, considering and selecting what features of the situation to attend to, remembering relevant information, interpreting perceptual and verbal information, seeking further information and making revisions, drawing inferences, predicting outcomes, and accepting or endorsing relevant beliefs, attitudes, values, obligations, and responsibilities. A person’s deliberative process results in a judgment about what they will do (a decision) and an intention to try to do it (see Bratman [2009] for a prominent account of intentions). To have decisional autonomy is to be able to complete one’s deliberative process without being hindered by factors internal to one’s cognitive machinery or by factors external to it. An inability to make decisions would make self-governance impossible. Although the decision-making process and decisional autonomy are aimed at (outward) actions and outcomes, they are relatively “inward” because they do not yet involve efforts to realize one’s intention in acting.

In addition to making decisions, inward autonomy involves authenticity, which refers to a person’s ability to recognize their choices and actions as “their own” (Christman, 2009; Friedman, 2003; Meyers, 2005). Authenticity involves critically reflecting on, evaluating, and authorizing or endorsing one’s choices and actions (Stoljar, 2024). The philosophical literature characterizes authenticity in various ways. To be authentic, one’s choices must cohere with one’s character (Dworkin, 1994), or with one’s stable preferences and deeply held values (Jaworska, 1999), long-term plans (Bratman, 2018), life-projects (Dworkin, 1988), most thoroughly integrated psychological states (Arpaly & Schroeder, 1999), or second-order volitions—i.e., the desire that one particular desire rather than another will move one to action (Frankfurt, 1988). The idea of authenticity presupposes that certain aspects of the self are more truly self than others. It captures the intuition that we are not autonomous simply because we do something we want to do; why we want to do it, and how we formed our judgment and intention to do it also matter. We can be moved to choose and act by desires that do not reflect the sort of person we wish to be, as with Frankfurt’s famous case of the unwilling drug addict who desires drugs and acts in accordance with this desire but does not sanction or condone their action (they wish they were not addicted to drugs). Their action is not autonomous because, although they perform it, they cannot identify with or endorse the action from the standpoint of their authentic or reflective self. Similarly, a person coerced or manipulated into performing actions they would not reflectively endorse has had their autonomy undermined even though they decided to act in those ways and initiated those actions. Inward autonomy can be compromised or undermined by things that others do and by oppressive social conditions more generally.

Our account so far is compatible with purely procedural (or “formal”/“content-neutral”) views of autonomy. Procedural views make authenticity a matter of mere coherence between a person’s motivations for acting and their reflective point of view about the action (Dworkin, 1988; Frankfurt, 1988), without requiring that the agent’s point of view be reasonable or sensitive to relevant objective facts. Such accounts promote diversity, help to resist paternalistic intrusions, and reflect the intuitive idea that one can exercise self-governance without being morally correct and even without acting according to one’s own best interests. But many theorists feel that authenticity requires more than coherence with a point of view, however poorly formed. In addition, it is thought that the agent’s point of view must at least follow recognizable norms of reasoning (weakly substantive accounts) or must be responsive to the objectively determinable reasons that exist for holding their point of view (strongly substantive accounts) (Buss & Westlund, 2018). Some weakly substantive accounts make the possession of certain attitudes toward oneself (self-respect, self-confidence, self-trust, a sense of self-worth) a condition for autonomy, since without such attitudes, an agent will be unable to exercise the authority necessary to govern themselves (Govier, 1993). Relatedly, certain emotional states (e.g., intense shame or anxiety) can undermine the possibility of autonomous agency (Brasseur et al., 2013). Substantive accounts of personal autonomy, unlike procedural accounts, capture the intuition that someone who internalizes oppressive attitudes (e.g., false stereotypes) and makes these attitudes “their own” is not truly autonomous (Stoljar, 2024).

Rather than taking a stand on the debate between procedural and substantive accounts, we offer two lessons for those interested in studying or supporting personal autonomy in persons living with dementia. First, caution should be exercised when introducing substantive normative constraints as requirements on autonomy, since doing so may disregard diverse conceptions of the moral good and function as an excuse for unwarranted paternalism. Second, the recognition of at least some substantive aspects of autonomy (e.g., norms concerning self-trust and self-worth) may be necessary to detect ways in which persons living with dementia may have internalized autonomy-undermining attitudes that impede the exercise of authentic personal autonomy.

Outward autonomy

Inward autonomy, consisting of decisional autonomy and authenticity, constitutes the first set of abilities necessary for being able to do what one really wants to do. But forming authentic decisions only takes us part of the way; personal autonomy also requires the outward-facing ability to translate our decisions into actions in the world around us. Dividing autonomy into inward and outward aspects follows a distinction first drawn by Collopy (1988) between “decisional” and “executional” autonomy, with what we are calling “outward” autonomy essentially elaborating the meaning of “executional” autonomy in Collopy’s distinction. Outward autonomy, like inward autonomy, consists of a set of capacities, but this time involves capacities oriented more directly toward physical action, like communication and motor control skills. To realize many of our goals, we require the involvement of others; therefore, all the skills that go into effective communication are relevant to outward autonomy, from basic verbal and non-verbal language skills to skills of presentation and persuasion, awareness of one’s audience, empathy, emotional regulation, etc. To effectively translate our decisions into the actions that will bring about our intended aims or outcomes, we also need to exercise control over our own bodily movements and, through these movements, over various objects and tools in our environment. We need practical reasoning abilities to select appropriate means for achieving our goals and to construct strategies and formulate realistic plans that have a good chance of succeeding. To do what we truly want to do also depends on external circumstances lining up in fortuitous ways. So, practical abilities that enable one to bring about change in one’s circumstances are necessary. Many social, economic, and political factors support or hinder the autonomy of individuals. Money may not be able to buy love, but it certainly helps procure what is necessary to achieve many of our goals. Poverty, by contrast, shrinks a person’s options, preventing them from attaining many goals. So outward autonomy can be said to depend at least partly on luck or providence.

Many theorists, including prominent feminist scholars, have presented relational accounts of autonomy that emphasize the way the self is embedded in social relations, by contrast to accounts inspired by Immanuel Kant or John Rawls that involve an “atomistic,” self-sufficient view of the self (Stoljar, 2024). Relational accounts seek to do justice to a richer, more realistic account of personal autonomy “as a characteristic of agents who are emotional, embodied, desiring, creative, and feeling, as well as rational, creatures; and they highlight the ways in which agents are both psychically internally differentiated and socially differentiated from others” (Mackenzie & Stoljar, 2000, p. 21). They give attention to “the specific ways in which oppressive socialization and oppressive social relationships can impede autonomous agency” (Mackenzie & Stoljar, 2000, p. 22). Conceptions of “relational autonomy” differ on whether relationships are merely external conditions necessary for personal autonomy (causally relational accounts, as in Christman [2004]), or whether, more strongly, relationships with others partially constitute an individual’s personal autonomy (constitutively relational accounts, such as that found in Westlund [2009, 2012]). Relatedly, theorists question whether the “self” in individual self-governance is best conceived as inclusive or exclusive of others (Anderson et al., 2025). Personal autonomy consists of a set of capacities attributable to the individual person, but we must acknowledge the profound ways in which the development and maintenance of autonomy depend on social interactions and conditions. Some have suggested that, at least for some non-Western cultures, a concept of “family autonomy” might be more appropriate than the notion of individual autonomy (Helwig, 2006; Sedig, 2016). Clearly, a person’s ability to exercise autonomy can be affected by others, as when disparaging attitudes undermine a person’s self-confidence or when social conditions afford the individual no meaningful options. But should the social relationships and societal conditions necessary for the realization of personal autonomy be considered as part of the definition of autonomy itself? If a person’s sense of self-worth is a core component of personal autonomy, and self-worth inherently involves relations to others, it would seem that such relations partially constitute personal autonomy (Benson, 1994).

We take a neutral stance in relation to disputes between causally and constitutively relational accounts. It seems obvious, however, that relationships with others are profoundly important for the development, maintenance, and enactment of personal autonomy. Neither inward nor outward autonomy can grow or flourish except in the soil of supportive social contexts (Agich, 2003; Stumpf et al., 2022). Accordingly, any account of autonomy that hopes to accurately represent autonomy as it really exists in human beings must be relational in at least the causal sense.

Figure 1 summarizes the components of personal autonomy. Note that the specific capacities represented by the boxes at the bottom of the diagram are illustrative and are not meant to be exhaustive. Each of the two sets of capacities denoted by inward and outward autonomy is necessary for fully-fledged personal autonomy, conceived as the ability to do what one truly wants to do. It seems unlikely that we ever attain perfect autonomy in the sense of being able consistently to form and execute decisions in complete harmony with the point of view of our most authentic selves. Each of us is autonomous to a greater or lesser extent, depending on how developed our capacities are in the array of areas relevant to personal autonomy. Each of these capacities can be hindered by factors related to our upbringing and development, our social, political, and economic conditions, our own poor choices, the choices and actions of others, and our genetic makeup. Given all this, personal autonomy appears to be a delicate and difficult thing for any of us to sustain. But this is all the more so for persons living with dementia, who experience unique challenges that threaten to severely curtail or completely undermine autonomy.

Figure 1.

Concept-map of personal autonomy.

Part 2: testing the concept-map: lived experience of persons living with dementia

We now turn to the empirical literature to test the concept map in relation to the lived realities of persons living with dementia. To accomplish this, we completed a scoping review of literature attending to the lived experience of autonomy for persons with dementia. For a detailed methodological description, see online supplementary material Appendix A, and for a list of papers evaluated for full-text extraction, see online supplementary material Appendix B. After screening, 82 studies were considered eligible for inclusion. For a PRISMA diagram, see Figure 2. The following sections present the findings of the scoping review in relation to decisional autonomy, authenticity, executional autonomy, and relational embeddedness of autonomy.

Figure 2.

PRISMA diagram for empirical literature review. This diagram was generated using Covidence’s PRISMA diagram feature.

Decisional autonomy and dementia

The empirical literature confirms that decisional autonomy, the capacity to deliberate and form intentions about what one wishes to do, remains important for persons living with dementia despite diminished cognitive abilities. A large proportion of the studies in our review (56/82 or 68%) investigated decisional autonomy, examining decision-making processes carried out independently or collaboratively with others. The decisions spanned various domains, including daily routines (Fetherstonhaugh et al., 2013; Keyes et al., 2019), medical care (Dening et al., 2017; Kinch et al., 2024; McWilliams et al., 2018), living arrangements (Auriemma et al., 2024; Garvelink et al., 2019), driving (Adler, 2010; Stevenson et al., 2019), and financial management (Boyle, 2013a). Several studies showed that persons living with dementia strongly valued participating in decisions (Davis et al., 2017; Lai et al., 2019; Miller et al., 2016; Whitlatch & Menne, 2009), with one person in Fetherstonhaugh et al.’s (2013) study expressing the sentiment, “even though I know probably now I’m not contributing a great deal, at least I feel as if I’m part of the decision. And that’s very, very important” (p. 148). Likewise, participants in Wolfe et al.’s (2021) study noted that “the most important thing… was being involved in decision-making” (p. 1882). The literature attests that decision-making is a key part of the experience of autonomy for persons living with dementia. Evidence presented below will demonstrate the ways in which decisional capacities are expressed, achieved, and sustained in the unique context of dementia.

In the early stages of dementia, decision-making does not differ substantially from that of people not living with dementia. Dixon et al. (2021) highlighted that people with dementia are social actors “able to establish priorities and make decisions about their daily activities” (p. 2). Similarly, Black et al. (2013) observed that individuals with mild cognitive impairment “can and do engage to some extent in the decision-making process” and can articulate their preferences for future decision-making (pp. 1, 7). Like people without cognitive impairment, those with mild dementia engage in deliberative processes, weigh options, and provide rationales for their choices that reflect their values and preferences. With increasing symptoms of dementia, individuals typically retain more autonomy in everyday decisions than in complex decisions about medical treatments or living arrangements, as these choices have higher stakes and more profound consequences (Bonds Johnson et al., 2021; Smebye et al., 2012). Furthermore, decision-making styles evolve as dementia progresses. Bhatt et al. (2020) documented progression from “managed autonomy,” where carers facilitate decisions through supportive strategies like discussion and negotiation, to “delegated decision-making,” where persons living with dementia consciously transfer decisional responsibility to a trusted party. The transition between these styles is typically gradual rather than abrupt, with Samsi and Manthorpe (2013) observing a shift from active supported decision-making to more passive substituted decision-making over time. Delegating decisions to trusted others is a way to continue to exercise personal autonomy, in the sense of doing what one really wants to do, so long as the representative mirrors the individual’s own deliberative process.

The empirical literature suggests that persons living with dementia retain capacity for decisional autonomy when appropriately supported. For instance, Feinberg and Whitlatch (2002) found that 90% of care receivers with cognitive impairment could successfully identify preferred decision-makers and articulate everyday care preferences, such as what to wear or what to do. Furthermore, Horton-Deutsch et al. (2007) documented that “older adults with mild to moderate dementia can participate in the [health care decision-making] process, make choices, and provide reasonable explanations for their choices,” affirming the potential for reasoned deliberation (p. 105). However, increasing disorientation with regard to time and place accompanying disease progression can result in decisions that do not reflect the individual’s best interests and do not cohere with objective norms of reasoning, particularly in situations involving safety and risk (Adler, 2010; Jennings et al., 2017; Smebye et al., 2016; Stevenson et al., 2019; van der Weide et al., 2025). Additionally, the extent to which decisional autonomy persists in practice is highly dependent on external factors. Factors that enhance autonomy include allocating sufficient time for reflection, as persons living with dementia often require more time than typically provided in healthcare settings (Tyrrell et al., 2006), and employing supportive communication strategies that facilitate expression. Conversely, decisional autonomy is significantly hindered by practices such as premature exclusion, as seen by Kelley et al.’s (2021) observation that healthcare professionals frequently bypass or “override over the person [living with dementia] and go straight to ask the family,” even when involvement is possible (p. 962). Decisional autonomy in persons living with dementia is not merely a function of intrinsic cognitive ability but is profoundly influenced by social and environmental supports or barriers.

Authenticity and dementia

The empirical literature also confirms that authenticity, which refers to a person’s ability to recognize their choices and actions as “their own,” remains relevant and important for persons living with dementia despite cognitive decline; 30 of the 82 studies in our review (37%) included at least one aspect of authenticity as a major focus. As outlined in our concept map, authenticity involves critically reflecting on, evaluating, and endorsing one’s choices in relation to one’s stable preferences, deeply held values, and sense of self. While authenticity was less frequently discussed compared to decisional autonomy, many studies examined concepts related to authenticity, such as “sense of self,” “identity,” and “personhood,” with the literature often associating authenticity with continuity of self. Studies found that maintaining a connection to one’s identity and values was cited as important by persons living with dementia themselves. For instance, Fetherstonhaugh et al. (2013) highlighted that the essence of decision-making for persons living with dementia was feeling “I am still here!,” linking their decisions to their long-standing sense of self. Sinclair et al. (2019) documented participants expressing the need for decisions to align with their core selves, such as one individual wanting choices to reflect “what the real me… wants,” and another wishing to be “acknowledged and recognized that you’re still a person and you’ve still got the ability to reflect what it is that I want” (pp. 594–595). The literature suggests that persons living with dementia greatly value having their authentic selves acknowledged. Evidence presented below will demonstrate the ways in which authenticity is expressed and sustained in the unique context of dementia.

Several studies associated authenticity with enduring identity and the extent to which a stable sense of self is retained despite the physical, cognitive, and personality changes that occur with dementia. Hedman et al. (2019) found that “past abilities contribute significantly to the sense of self of people with dementia,” with persons living with dementia being “more explicit about their agency in the past than were the carers” in the sense that they highlighted their previous accomplishments and capabilities as central to their identity more so than their carers did when describing them (p. 1366). Basting (2001) found that even when individuals lose the ability to weave a traditional narrative of selfhood from memory, engaging in creative storytelling allows them to shape their story with the fragments of memory remaining. The study identified “nuances of individual personalities” present even within fragmented personal narratives and helped participants reinforce a sense of authentic self, allowing for continued autonomy. In the context of advancing symptoms of dementia, authenticity is increasingly expressed through embodied and habitual actions rather than explicit verbal articulation. Some studies demonstrated that affective aspects of the self persist even when declarative memory deteriorates. Francis et al. (2020) showed that despite significant memory loss, a participant “still reacted emotionally like the same person [they] had always been” (p. 154). The consistency of individuals’ emotional reactions, expressed as patterns developed and replicated throughout their lives, was seen as pointing to an enduring aspect of self. Similarly, Rushton et al. (2024) observed that embodied and habituated actions can indicate consistent desires and preferences and represent authentic expressions of agency. Importantly, these embodied and habituated actions all occurred even when conscious recognition or explicit verbal articulation was diminished.

A second conceptualization, much less emphasized in the literature, understands authenticity as a reflective process that can occur in the present moment without necessarily referencing prior preferences. This perspective acknowledges that values and preferences can evolve as one adapts to changing circumstances while still remaining authentic expressions of the person’s current self. This is exemplified in Davies et al.’s (2021) study, where a person with dementia articulated clear reasoning about future living arrangements: “At what point do I give up living where I am? I think it’s when I become bedridden, that’s one. Two—when I don’t know what I’m doing. And three—if I’m in a lot of discomfort and pain or something like that” (p. 8). Statements like this demonstrate how persons living with dementia can and do engage in authentic reasoning about significant future decisions by reflecting on current values (e.g., quality of life, awareness, comfort) and establishing specific, personal criteria for a major life change. Jennings et al. (2017) found that people living with dementia and their caregivers recognized that preferences and goals are not static, with participants articulating “the need to re-evaluate goals over time as the needs of the person with dementia changed with progression of the disease” (p. 5). This suggests that authenticity can be rooted in a reflective process that refers to present values that may differ from those previously held and need not be consistent with previous preferences.

The empirical literature suggests that authenticity can persist substantially in persons living with dementia when appropriately supported, though the mechanisms for its expression evolve with disease progression. Supporting factors primarily involve caregivers' knowledge of and respect for the person’s history and values. Leverton et al. (2021) highlighted that being able to connect past identity with clients' present lives and choices was important for homecare workers providing autonomy-supportive care. Family members also play a role in maintaining authenticity, with Samsi and Manthorpe (2013) finding that many persons living with dementia trust their partners to make decisions aligned with their authentic selves, as the caregiver “most often tended to fall back on their accumulated familiarity with and knowledge of their partner, including their preferences, habits, and beliefs” (p. 956). Toms et al. (2015) found that for people with dementia, maintaining “old ways of coping” and asserting their “retained abilities” contributed to their sense of continuing identity and self. Family carers supported these efforts, for instance, by encouraging consistent routines (Toms et al., 2015). Authenticity can also be hindered, however, when proxies fail to distinguish between the person’s authentic wishes and their own preferences or concerns. Horton-Deutsch et al. (2007) found that in decisions about hypothetical hip fracture treatment, caregivers often prioritized safety concerns, opting for physical therapy for the person with dementia, even when the individuals themselves expressed a preference for surgery and reasonably articulated their justification (to regain function and activity more quickly). Such substituted decisions, while often made with protective intentions, can undermine the authenticity of the person living with dementia, because the imposed choice no longer reflects their reflectively endorsed values and preferences.

Executional autonomy and dementia

The empirical literature confirms that persons living with dementia highly value executional autonomy, which involves the ability to translate decisions into actions in the world. Thirty-five out of the 82 studies (43%) involved a major focus on some aspect of executional autonomy. As outlined in our concept map, executional autonomy encompasses effective communication skills, control over bodily movements, practical reasoning abilities, and the capacity to bring about changes in one’s circumstances. Multiple studies in our review examined how persons living with dementia enact their choices and intentions through physical action. We noted the following key terms: (1) “Independence,” which involves choice, control, the ability to perform daily activities, freedom of movement, opportunities to act for oneself, and making decisions about one’s time (Evans et al., 2007); (2) “Embodied agency,” the expression of agency through non-verbal physical actions and body language, particularly when deliberative or verbal capacities are limited (Boyle, 2014); (3) “Implicit memory,” the capacity to recall procedures without conscious effort (González-Cabañe et al., 2022), which enables persons living with dementia to translate intentions into physical actions and continue performing familiar motor tasks and routines through automatized processes, even as “declarative memory,” the conscious recollection of particular facts, diminishes; and (4) “functional independence,” the maintaining of physical functioning, such as performing self-care and household tasks (Jennings et al., 2017). The literature on the lived experience of autonomy of persons with dementia shows a significant interest not only in decision-making but also in action implementation, which is the heart of executional autonomy.

In the context of dementia, a progressive shift takes place from cognitively mediated to more embodied forms of action implementation. Implicit memory often persists longer than declarative memory (González-Cabañe et al., 2022). One of the studies in Boumans et al. (2019) illustrates this persistence of implicit memory in supporting action with regard to an individual’s underlying capacity to enact a familiar, multi-step gardening routine. The individual’s sense of gardening being intrinsically “part of his life” points to the significant role of procedural (implicit) memory in sustaining this form of executional autonomy. The retention of embodied knowledge represents a distinct pathway for expression of executional autonomy as dementia progresses; decisions and preferences rooted in the individual’s character are expressed in action without conscious deliberation. González-Cabañe et al. (2022) showed how participating in artistic activities enabled increasing independence over time through repeated engagement, demonstrating enhanced executional autonomy by requiring less direct assistance to perform the necessary sequence of physical actions involved in painting. Improvements were most pronounced among those with severe dementia. This suggests that autonomy can be maintained and even improved through routine physical activities even as cognitive functions diminish. Especially in later stages when verbal abilities may decline, executional autonomy is increasingly conveyed through non-verbal expressions. These physical actions, gestures, and behaviors become crucial for persons living with dementia to continue exerting their will and preferences. Motta-Ochoa et al. (2022) observed persons living with dementia participating in an arts-based movement program maintaining agency through embodied expressions, understood as the observable, physical ways individuals enact their intentions by modifying or adapting the program’s suggested physical exercises to better fit their personal interests and desires, resisting those that did not align with preferences, and improvising movements collectively. These expressions, along with the skillful performance of learned implicit tasks like painting (González-Cabañe et al., 2022), represent an evolving manifestation of executional autonomy.

The empirical literature indicates that executional autonomy can persist substantially in persons living with dementia when appropriately supported, though the extent varies considerably based on disease progression and environmental factors. In both decisional and executional autonomy, environmental context plays an important role. Rapaport et al. (2020) identified “being in a safe familiar environment” as key to maintaining independence at home. In this practical sense, “independence” is closely aligned with executional autonomy, as it reflects the capacity to implement choices and manage daily living through effective physical action. Zhang et al. (2025) demonstrated how environmental design modifications that use visual cues, lighting, and simplified tools can enhance spatial orientation and object interaction and subsequently support executional autonomy. Boumans et al. (2019) found that “specially designed spaces… provide residents with an activity that reminds them of who they were or what they loved in the past; this in turn can make them feel more comfortable and confident” (p. 725). Authenticity and executional autonomy interact here as the increased comfort and confidence stemming from this connection to self bolsters capacity and motivation to perform these familiar actions. With regard to decisions about living space (and other topics), executional autonomy can also be undermined in the name of ensuring safety. Jennings et al. (2017) identified conflicts between caregivers' goals to ensure safety and the goal of persons living with dementia to maintain functional independence.

Meaningful executional autonomy can persist even amid significant cognitive and physical impairment when supported through accessible environments and recognition of non-verbal expressions of agency. This understanding challenges deterministic views of dementia progression that portray dementia as an inherently linear and irreversible decline leading to the inevitable and complete loss of personal autonomy.

Relational embeddedness of autonomy and dementia

The empirical literature strongly confirms that autonomy is embedded in and enabled through relationships with others for persons living with dementia. Seventy of the 82 (85%) included studies contained a major focus on some aspect of the relational embeddedness of autonomy. Notably, a significant proportion (n = 54) of the 66 original studies in our review included not only persons living with dementia but also their family carers or professional supporters, often through dyadic or multi-perspective methodologies. This common research approach itself underscores an implicit understanding within the field that the development and exercise of autonomy in persons living with dementia necessarily requires relational processes, and even lends some support to the stronger view that such processes partly constitute personal autonomy for members of this population. The socio-relational embeddedness of personal autonomy is conveyed in frequently used terms such as “interdependence,” which refers to mutually beneficial relationships where autonomy is maintained through shared support. Keyes et al. (2019) notably describe interdependence as a negotiated and mutual state that participants explicitly value over isolated independence or complete dependence. Miller et al. (2016) describe “shared decision-making” as collaborative processes involving multiple stakeholders in reaching decisions, and “advocacy” as the representation and promotion of a person’s interests by trusted others when communication barriers exist. Many studies in our review examined how relationships enable or constrain the exercise of personal autonomy, reflecting the centrality of social and relational context for understanding autonomy in persons living with dementia.

The relational embeddedness of personal autonomy manifests in the increasingly interdependent processes that emerge as cognitive abilities change. Autonomy in persons living with dementia progressively shifts from independent self-governance to supported expressions of agency through trusted relationships. Groen van de Ven et al. (2017) demonstrate this shift through their analysis of decision-making processes in dementia care networks, showing how decision-making becomes interdependent between people with dementia, informal caregivers, and professionals. This interdependence can appear through specific support mechanisms, such as informal caregivers actively initiating discussions about options and working to bridge conflicting viewpoints that may arise. Further illustrating this reliance on interdependence, Keyes et al. (2019) document how persons living with dementia navigate everyday decisions through negotiated relationships. Their work highlights the importance of what Fetherstonhaugh et al. (2013) termed “subtle support” rather than “taking over,” where care partners attentively interpret both verbal and embodied expressions to understand preferences and resist overriding the desires of the person with dementia. Interdependence becomes increasingly integral to autonomy expression as dementia progresses, requiring strategies that adapt to cognitive decline while preserving aspects of personal autonomy (Miller et al., 2016).

The empirical literature indicates that personal autonomy can persist substantially in persons living with dementia when supportive relationship dynamics are present, with the form of support evolving along with disease progression. The main enhancing factors have to do with caregivers' communication approaches and attitudes. Boyle (2013b) identified multiple strategies spouse-carers use to facilitate decision-making, including various forms of supervision and support that preserve aspects of autonomy while providing necessary assistance. In healthcare contexts, McWilliams et al. (2020) observed that, in cancer–dementia dual diagnoses, carers play crucial roles in ensuring the person’s understanding and participation by, e.g., breaking down complex medical information and relaying it in accessible ways, thereby supporting the person’s ability to engage with decisions about their care. Sinclair et al. (2019) found that relationships based on trust significantly enhance autonomy for people with dementia. Furthermore, Lai et al. (2019) noted that healthcare professional relationships built on trust and regular contact were most conducive for discussions about future needs, since personal autonomy in persons with dementia is more sensitive to relational dynamics. However, this also means that personal autonomy can be hindered by power imbalances or communication breakdowns with supporters. Tyrrell et al. (2006) documented substantial disparities between how persons living with dementia and their caregivers perceived decision-making processes, finding that while 62% of caregivers felt they had been “very well listened to by professionals,” only 38% of persons living with dementia felt similarly heard, raising questions about whether “the carer’s views are more actively sought than those of the patient” (pp. 488, 490). Autonomy’s persistence depends not just on the presence of relationships but on their quality, characterized by trust, attentive listening, and balanced communication that preserves the centrality of the person living with dementia’s perspective even as external support increases.

Part 3: synthesis of theoretical and empirical findings

The goal of this paper is to present an account of personal autonomy that is theoretically and empirically informed and helpful both to researchers wishing to study autonomy in persons living with dementia and to healthcare providers wishing to support autonomy in this population. Part 1 derived from the relevant philosophical literature a concept map outlining the main aspects of personal autonomy. Accordingly, personal autonomy is the socially and relationally embedded capacity to make and execute authentic decisions or, more informally, the ability to do what one really wants to do. This account reflects a broad consensus within philosophical theorizing about personal autonomy despite contested issues within the field. Part 2 used the results of a scoping review of empirical studies of personal autonomy in persons with dementia to test the concept map from Part 1. We determined the prevalence in this literature of each main element of the concept map (decisional autonomy, authenticity, executional autonomy, and relational embeddedness) and elucidated patterns concerning how these aspects manifest in the context of living with dementia. This third part brings Parts 1 and 2 together to suggest what a theoretically and empirically informed framework for understanding personal autonomy in the context of dementia should involve.

Alignment between theory and observation

The three main components of personal autonomy represented in our concept map are all central concerns of the empirical literature on personal autonomy in persons living with dementia, with decisional autonomy and executional autonomy as a main focus (68% [n = 56] and 43% [n = 35], respectively) of the 82 studies included and authenticity as a main focus of 37% (n = 30) of studies. Furthermore, the socio-relational embeddedness of personal autonomy was a focus of 70 of the 82 studies (85%). Many studies purporting to study “autonomy” in fact focus only on one or another aspect of autonomy (e.g., decision-making), and few make a concerted effort to operationalize the term with reference to a conceptual model. (Indeed, one of our goals is to elucidate the whole concept of personal autonomy to reduce the ambiguity and incommensurability that affects the empirical literature on autonomy.) It is unsurprising, then, that we did not find a still greater proportion of studies representing the main elements of the concept map.

On the whole, the empirical literature we reviewed supports the claim that people living with dementia can exercise all aspects of personal autonomy to varying degrees. This is more obviously the case for individuals with mild to moderate dementia, but several studies indicated that even in late-stage or severe dementia, autonomy can be realized to a significant extent so long as appropriate supports are in place. The symptoms of dementia diminish the operation of the capacities necessary for personal autonomy but do not erase them. This finding aligns with Part 1’s depiction of the relevant capacities as “sliding scales” and the corresponding recognition that every human being, whether experiencing dementia or not, expresses autonomy at varying levels depending on a multiplicity of internal and external factors affecting these capacities. None of us is ever perfectly autonomous in the sense of having all necessary capacities operating at peak level. As theorists of relational autonomy have convincingly insisted, the development and exercise of every person’s autonomy depends profoundly on others and on various social, cultural, linguistic, economic, and technological systems. These conceptual points, alongside the corroborating empirical evidence, should move us to a sympathetic recognition of the capacities of those of us who must exercise our agency in the context of living with dementia.

In several ways, attending to the empirical literature helps to concretize our theoretical understanding of personal autonomy, which might otherwise remain blind to particularities of living with dementia and the way autonomy manifests in that context. First, this literature points to specific areas in which the exercise of autonomy becomes problematic for persons living with dementia. As symptoms emerge and progress, areas of decision and choice previously taken for granted—including activities of daily living, healthcare options, driving, finances, and living arrangements—require increasing attention to balancing freedom and risk. Second, the literature indicates numerous ways to support and preserve personal autonomy in light of diminishing cognitive capacities. For instance, personal autonomy can be enhanced by supportive communication strategies (e.g., simplifying information), use of adaptive decision-making styles, engaging in creative storytelling, helping to maintain consistent routines, noticing non-verbal modes of expression and the role of implicit memory, and modifying the physical environment to increase a sense of familiarity and confidence. Factors that hinder the expression of personal autonomy include premature exclusion from decision-making conversations, communication breakdowns, failure to allow enough time for processing, substituting carers' preferences or concerns for those of the individual, and placement in disorienting or unfamiliar environments. Third, the multifaceted socio-relational support that is needed for anyone to exercise personal autonomy (as per our concept map) comes to the foreground in concrete ways in view of the increasing dependence on others that accompanies progressing symptoms of dementia. The empirical literature reflects this in its concrete discussions of interdependence, shared decision-making, and supported agency. The more overtly socio-relationally supported autonomy of those living with dementia differs only in degree and not in kind from personal autonomy as it manifests in those who do not experience dementia symptoms. Without attention to such contextual details, it would be easy to conclude that a generic theoretical account of personal autonomy does not apply to individuals living with dementia, while in fact their capacities are merely suppressed by unexamined socio-relational factors.

Divergence between theory and observation

As we have seen, the concept map’s depiction of personal autonomy is strengthened by the concrete discussions in the empirical literature. At the same time, the concept map serves as a standard that can indicate incomplete or tendentious conceptualizations of the elements of personal autonomy in the empirical literature. Authenticity serves as the most evident example of this. Nearly all of the papers concerned with authenticity assumed that it can or should be assessed by determining whether a person’s present choices reflect or align with what is known about their prior preferences, beliefs, and values. In philosophical theorizing about personal autonomy, by contrast, the core of authenticity concerns the person’s reflective endorsement of their choices, which may take place by reference to values held previously, but may also take place “occurrently,” by reference to present values that differ from those previously held (Mackenzie & Stoljar, 2000). The assumption that a person’s “real self” must equate to their self prior to the effects of dementia is at least questionable. Continuity with one’s prior self would arguably be irrelevant for a person who had never been able to make authentic decisions because of oppressive socialization or ongoing oppressive relationships. Furthermore, Ronald Dworkin’s (1994) prioritization of precedent autonomy (the wishes of the competent person prior to advancing dementia) over contemporaneous autonomy of the person living with dementia has been critiqued by Rebecca Dresser (1995) and others, with implications for debates about the status of advance directives for persons with dementia who lose competence (see Lin et al., 2025). We do not have space to address this debate fully here, but we strongly caution against an overly simplistic identification of authenticity with consistency between present and prior values and interests.

At the risk of speculating, the close link in the empirical literature between authenticity and continuity with one’s prior self might be explained by the stereotype that as dementia progresses, personhood is gradually lost. Depictions of dementia in popular media are often framed as a “living death” and may use language such as “vegetable,” “there is nobody there,” and “withered shells” to describe those living with dementia (Low & Purwaningrum, 2020). These dehumanizing descriptions of people with dementia reflect a dominant societal narrative that the neurological changes associated with dementia inevitably lead to a loss of the “self” and one’s “true” identity. As a result, there may be a felt need to “prove” that personhood and the resulting capacity for authenticity remain by showing continuity with an earlier version of the self whose personhood was not in question. It is also possible that reflective endorsement is particularly difficult to assess in someone living with dementia whose ability to communicate has been affected. Whatever the reasons for it, discussions of authenticity in the empirical literature tend to reflect a narrow interpretation shaped by prevailing assumptions about personhood and dementia. Philosophical accounts of authenticity could offer a more expansive view to the empirical literature, recognizing that the ability to recognize one’s choices and actions as “their own” does not necessarily require consistency with one’s past actions.

The empirical literature diverges in a less blatant way from the philosophical in its frequent reduction of executional autonomy to independence, with independence itself often operationalized as an ability to live in one’s own residence rather than in assisted living or long-term care. Rapaport et al. (2020), for instance, thematically describes how people living with dementia maintain independence at home and resist overreliance on caregivers, reflecting the idea that the ultimate goal for people living with dementia, whether realized or not, is to “do it themselves.” The problem is that, understood as a set of capacities that enable us to translate authentic decisions into actions and outcomes, executional autonomy is very much compatible with dependence, since the focus is on whether the person can make happen the things they authentically want. A person can realize their decisions either directly, through their own physical actions, or indirectly, through the physical actions of other persons. Given that personal autonomy always depends profoundly on supportive social and relational contexts, the simple identification of autonomy with independence should be resisted. These two concepts should be more carefully distinguished, and their relationships should be made clear.

Similarly, although the increased concreteness of considering specific domains of decision-making is helpful, it can also be limiting when trying to achieve a broad understanding of decisional autonomy. Our concept map identifies decisional autonomy as a complex set of capacities needed to deliberate and form intentions concerning a course of action. But rather than investigating the experience of the decision-making process itself, many studies of decisional autonomy in persons living with dementia are driven by pragmatic questions about whether or not the individual was involved in, for instance, decisions about their healthcare.

Our empirical review generally confirms the socially and relationally embedded nature of personal autonomy introduced in Part 1. The fact that the design of most studies included persons living with dementia along with caregivers and healthcare professionals (55/66 original studies in our review), however, may reflect a potential bias. This design feature may unintentionally insinuate that “true” insight into the experience of the person with dementia will be more accurately reflected by their close contacts who are not experiencing dementia symptoms. Although this assumption makes possible the erasure of perspectives of persons living with dementia, given the cognitive and physical decline in dementia and the socio-relationally embedded nature of autonomy, it is reasonable to expect that such individuals may require relational support throughout the research process. Furthermore, caregiver support may be appreciated by persons living with dementia. Many empirical studies identified that a familiar environment was important for enhancing the autonomy of persons living with dementia (Leverton et al., 2021; Rapaport et al., 2020), and the presence of caregivers often represents a source of comfort during research participation. Though persons with dementia may benefit from caregiver involvement, it is important to ensure that their perspectives are held with equal weight to those of caregivers and providers. In joint interviews, care must be taken to ensure that the perspectives of persons living with dementia are not overridden by those of caregivers.

Conclusion

Nothing in the empirical literature suggests that the concept map from Part 1 is incomplete; no additional elements of personal autonomy were discerned in these studies beyond those identified in the map. We conclude that, at least as far as available evidence goes, this concept map is suitable as a basic framework for further study of personal autonomy in people living with dementia, and for healthcare policy aimed at supporting the autonomy of members of this population. Each of the core elements of the concept of personal autonomy, as derived from the relevant philosophical literature, remains applicable to persons with dementia, and with appropriate social and relational support, each element can be realized, at least partially. Each should therefore be included in attempts to understand the phenomenon of personal autonomy in persons living with dementia and in efforts to support it. Our findings also suggest, however, that when applying this generic conception of autonomy to the lived experience of dementia, it is worthwhile to consider several concrete features and nuances that characterize that specific context. Most importantly, given the reality of diminished capacities that can accompany the progression of dementia, extra attention must be given to ensuring the presence of social and relational dynamics known to preserve and enhance autonomy and to minimizing factors that commonly hinder its expression. We have attempted to balance and integrate philosophical/conceptual and empirical/social-scientific approaches to produce an account of personal autonomy that is both conceptually clear and empirically adequate. To the extent that we have succeeded, this account offers direction for more effective study of personal autonomy and more helpful autonomy-promoting interventions for persons living with dementia.

Supplementary material

Supplementary data are available at The Gerontologist online.

Funding

This work was supported by the Social Sciences and Humanities Research Council of Canada, Insight Grant [435-2022-0530 to A.D.H.S.].

Conflict of interest

None declared.

Data availability

The data underlying this article are available in the article and in its online supplementary material. The review protocol was registered with Open Science Framework (https://osf.io/fy5eq/). Any additional data are available upon reasonable request to the corresponding author.