Article Text

Download PDFPDF

The implausibility of response shifts in dementia patients
  1. Karin Rolanda Jongsma1,2,
  2. Mirjam A G Sprangers3,
  3. Suzanne van de Vathorst1
  1. 1Medical Ethics and Philosophy of Medicine, Erasmus University Medical Center, Rotterdam, The Netherlands
  2. 2Medical Ethics and History of Medicine, University Medical Center Georg-August University, Göttingen, Germany
  3. 3Department of Medical Psychology, Academic Medical Center/University of Amsterdam, Amsterdam, The Netherlands
  1. Correspondence to Karin Rolanda Jongsma, Medical Ethics and Philosophy of Medicine, Erasmus University Medical Center, Rotterdam 3000 CA, The Netherlands; karin.jongsma{at}


Dementia patients may express wishes that do not conform to or contradict earlier expressed preferences. Our understanding of the difference between their prior preferences and current wishes has important consequences for the way we deal with advance directives. Some bioethicists and gerontologists have argued that dementia patients change because they undergo a ‘response shift’. In this paper we question this assumption. We will show that proponents of the response shift use the term imprecisely and that response shift is not the right model to explain what happens to dementia patients. We propose a different explanation for the changed wishes of dementia patients and conclude that advance directives of dementia patients cannot be simply put aside.

  • Dementia
  • Decision-making
  • Clinical Ethics
  • Living Wills/Advance Directives
  • Quality/Value of Life/Personhood

Statistics from


Frank has been suffering from Alzheimer's disease for at least 5 years. He barely reacts to questions asked by his physician and is only interested in sleeping and smoking cigarettes. For a couple of weeks he has been living in a special nursing facility. Overall, he does not seem to be unhappy, but he cries each time his wife Claire leaves him. Claire tells the physician that he acts completely differently from the way he used to. Frank was a very successful and clever politician who took full control of his life. He was passionate about gaining power and strove for success in his career, and overall lived a very independent life. Frank never wanted to be dependent and despised the prospect of being seen as a vulnerable and powerless person. Earlier in the disease process, the ‘old’ Frank would come back for a couple of minutes – he would throw away his medication, dress himself properly and boss everyone around. However, that has not happened for some time and currently he is always in a demented state. Claire doubts whether it was the right decision for Frank to stay in the nursing facility, and expresses her problems with this situation to the physician: ‘He would have never wanted it and would prevent being in this situation now if he still could’. The physician points out that Frank does not show any behaviour that confirms his prior objections, and that his quality of life (QoL) has not visibly changed since he started living in the nursing home, so there is no reason to worry. When they ask Frank how he feels about his stay in the facility, he nods and smiles.

One may wonder why we should consider Frank's prior preferences at all if he does not seem to mind living in a nursing home. Some argue that prior preferences do not matter anymore for dementia patients1–3 because dementia is a gradually progressive condition and psychological coping strategies will contribute to ‘coming to terms’ with the disease.1 Furthermore, so it is argued, what patients find important may change as dementia progresses; what was important in the early stages may seem unimportant in the late stages.4 ,5 Because this adaptation process is regarded by some as a ‘response shift’, we should not hold dementia patients to their previously expressed preferences.3 ,4 This view has far-reaching implications, namely that we can disregard prior preferences even when expressed in advance directives.

The aim of this paper is to show the flaws in understanding conflicting preferences and wishes of dementia patients as a response shift. The main questions we will answer are: (1) What is a response shift? (2) How is the term ‘response shift’ used for dementia patients? (3) Does the term response shift accurately describe the experiences of patients with dementia?

We will argue that dementia patients lack the capacities required to measure a response shift and that response shifts are an implausible explanation for the altered preferences of dementia patients.

What is response shift?

The concept of ‘response shift’ has its foundation in research on educational training interventions6 and organisational change.7 The concept refers to a change in the meaning of one's self-evaluation and may occur in any field where self-reports are required.6 Sprangers and Schwartz applied this concept to self-report in a medical setting, more specifically in the field of QoL research.8 QoL is a multidimensional concept reflecting a broad range of life domains from both a subjective and an objective perspective.

Response shift is commonly defined as the change in self-evaluation of QoL as a result of a change in (1) the respondent’s internal standards of measurement (ie, recalibration), (2) the respondent's values (ie, changes in importance or priority) or (3) the meaning of QoL (ie, reconceptualisation) by the respondent.8 The response shift model holds that the evaluation of QoL can change without alteration in health status, or that the QoL evaluation can remain stable in the face of significant changes in health status.8 This second type of response shift is often seen in patients with chronic diseases who change their self-evaluation of QoL as the disease progresses. To illustrate how certain changes may be associated with the process of accommodating the illness, Sprangers and Schwartz tell the following story in which an orthopaedic patient undergoes a response shift:A woman who, after hearing her diagnosis of osteosarcoma, told him [the physician] that if her bone tumor prevented her from being able to walk, life would no longer be meaningful to her and she would prefer euthanasia. When the time came that this woman was confined to a wheelchair, she informed him that life still held value for her but that if she were to become incontinent or bedridden, then life would lose its meaning and she would prefer euthanasia. However, when the time came that she was incontinent and bedridden, the woman stated vehemently that life still held meaning for her and that she was not ready for euthanasia. (ref. 8, pp. 1507–1508)

This story shows that self-evaluation of QoL can change over the course of a disease's trajectory and that these changes may be associated with the process of accommodating the illness. Changes in the patient's health may prompt behavioural, cognitive and affective processes necessary for accommodating the illness. These processes can change an individual's standards, values or conceptualisation of QoL, and a response shift may influence one's perceived QoL, and thus change one's preferences.

How is response shift used in dementia patients?

Some authors argue that people with dementia respond in the same way as people who are confronted with other illnesses. Dementia patients also try to actively cope with their illness and to make sense of their fractured world.3 ,9 It is argued that the gradual progression of the disease allows dementia patients to cope with and adapt to the disease,1 ,10 and that this explains why dementia patients change their preferences later in the disease process:In order to come to terms with dementia people use a variety of coping mechanisms … Dementia is not a disease that appears overnight, overwhelming the person involved. Instead, it is often a process of insidious onset and gradual progression, allowing the person to adapt and adjust to their changing situation. (ref. 1 pp. 1032–1033)

Adjustment and adaptation are seen as important psychological coping strategies that contribute to coming to terms with dementia.1 ,11 These coping mechanisms and adaptations to the disease are considered as evidence for response shift,5 ,11 ,12 or adaptation may be seen as being a synonym for response shift:10 ,11One of the causes of RS [response shift] is adaptation to illness. (ref. 12, p. 174)In  … dementia research … some research has shown that PLwD [people living with dementia] and their family caregivers may adapt (response shift) to the limitations of the disease in their daily life and that this may help to explain their positive evaluation of QoL. (ref. 10, p. 104)

Besides adjustment and adaptation, it is argued that dementia patients change their judgement of QoL during the disease process. In this view, it is suggested that dementia patients change their mind about what is important to them and set new priorities:4 ,5Judgments about what is important to quality of life may change as the dementia progresses: what seems important in early stages (eg, preservation of intellectual capacity) may seem unimportant in late stages, when safety and comfort may take on primary importance. (ref. 4, p. 2356)

These authors see response shift as explaining the contradiction between previously expressed preferences and the current wishes of dementia patients. Prospects that were feared or considered undesirable at first, such as admittance to a nursing home, might be acceptable to the patient in a later phase of their disease:9 ,11[A]dmittance to a nursing home may in itself also represent response shift and acceptance instead of suffering, which can be investigated by evaluating an AD [advance directive], if present, and by the patient's recent responses to the disease process. (ref. 11, p. 91)[I]n practice this will mean that patients will regularly contradict their own advance directive or—as is a very common clinical experience—adapt. (ref. 9 p. 50)

Explaining the phenomenon that dementia patients ‘change their mind’ as a response shift has benefits: it provides an explanation for the paradoxical finding that dementia patients living in a nursing home experience a relatively high QoL.13 And it provides an explanation and solution to the difficult situation when prior preferences conflict with current wishes. If the response shift approach were true, the patient's current wishes should be followed, as we would also do with patients with other illnesses. The prior preferences of the patient become irrelevant and the advance directives made by the patient can thus be ignored.

The impossibility of measuring response shift in dementia patients

However, the response-shift view for dementia patients is problematic in several ways. First and foremost, a response shift requires that patients are able to evaluate and report their QoL. Dementia patients progressively lose higher cognitive functions, especially functions such as memory, reasoning, comprehension, judgment and understanding. This results in impairments in concentration, logical reasoning, decision-making and communication.14 ,15 Dementia patients, especially those in advanced stages of the disease, are cognitively unable to provide self-reports of their QoL. Second, there are problems with QoL measurements in dementia patients regarding reliability (self-report is problematic due to declining memory and anosognosia or unawareness of deficits), validity (non-validated scales and decreased understanding due to declining cognitive capacities) and responsiveness (unclear what the outcome unit of measurement should be).10 ,12 Because of these problems, other approaches for measuring QoL in the more advanced stages of dementia have been developed, for example through observation or by proxy rating.16 ,17 These measurements are clinically relevant but cannot be used to demonstrate a response shift because that requires self-evaluation. Due to these methodological problems, empirical studies do not provide reliable evidence. Indeed, empirical studies show mixed results concerning the stability of QoL throughout the disease process. Some studies show a decrease in QoL,18 ,19 especially in the early stages of the disease when patients are still able to self-report, while other studies show that longitudinal measurements by observation or proxy rating in dementia patients do not show significant changes in QoL over time, even though the patient's cognition declines.13

Some authors argue that dementia patients undergo a response shift, but QoL measurements for dementia have important methodological challenges and dementia patients are not able to provide self-evaluation. Therefore, we cannot acquire evidence about response shift in dementia patients.

Why response shift is implausible in dementia

Beside the methodological challenges, it is also very unlikely that dementia patients change their internal standards, their values or their conceptualisations of QoL. Therefore, a response shift is an implausible and misleading explanation for the changing wishes of dementia patients.

Many of the proponents suggest that dementia patients adapt or adjust to the disease.1 ,3 ,5 ,10 Essential to a response shift is that the patients' internal standards, values or conceptualisation of QoL change. ‘Response shift’ is thus not a synonym for any change or adjustment. Here we should make an important distinction between preferences, values and wishes. Preferences are based on and guided by values. Values are developed and learned over time, adjusted, reflected upon and embodied, and form the foundation for decision making. Wishes may be unreflected and are not guided by values. The proponents of the response-shift view, therefore erroneously see the formulation of new wishes as adaptation and as evidence for a response shift. Furthermore, most of the proponents do not make explicit which of the three types of change (in internal standards, values or conceptualisation) they refer to when they discuss adjustment or coming to terms with the disease process. Byrne-Davis et al5 are an exception, as they argue that adaptation is due to a change in internal standards, but they do not explain why and how a change in internal standards would apply to dementia patients.

A change in terms of internal standards is often seen in patients with severe pain.20 While a patient might have said yesterday that his pain level was 10 out of 10 on the pain scale, the worst pain imaginable, today he might say: ‘I was wrong, yesterday's pain was bad but compared to today's pain that was only 8 out of 10, today is really the worst pain possible’. The standard with pain which is measured has thus been altered. In Frank's case, an example of changing internal standards would be if he reasoned: ‘Even though I thought living in this nursing home would be the worst that could happen to me, I have foreseen that wrongly as I now think that living in another nursing home would be even worse’. From the mere observation that he nods in reply to the question as to how he feels about his stay in the nursing home, we cannot conclude that he has changed his internal standards. Moreover, for a change in internal standards, a reflective stance is necessary: you need to remember your former position on the scale in order to be able to revise it; you need to remember your former pain and compare it to another experience of pain, in order to judge that your current pain is worse. Even though the change in internal standards does not have to be a conscious process,8 it requires reflective skills. Due to his cognitive decline it is unlikely that Frank is still able to contemplate his standards. Therefore, it is also unlikely that he will draw the conclusion that the standard he held on to, does not suit him anymore.

A change of values is equally unlikely for dementia patients. This would imply that a dementia patient reprioritises his values. For Frank it would mean that the values of independence and success lose their relative importance. We question whether that is indeed the case. Even if Frank might act like these values do not matter much to him anymore, we cannot conclude it is the result of a new set of priorities. Due to dementia, the capacities to remember and live according to prior preferences are lost. Some have argued (eg, Jaworska) that dementia patients remain able to value,21 but we question whether these remaining values should be considered leading values. Dementia patients do not choose what they forget, therefore what they still remember and the values they still express are not to be considered the ones with the highest priority. Which values and preferences remain is not the result of a conscious choice but is dictated by the disease. It is therefore a false conclusion to say that the dementia patient no longer ‘cares’ about the things that mattered when he did not have dementia.21 Not caring about something presupposes that the person could still care about it. The point here is not that he no longer cares, but that he is no longer capable of caring about his prior values—including the things that mattered to him before the onset of dementia.22

Finally, a change in the conceptualisation of QoL is also an unlikely explanation for the wishes of dementia patients that conflict with prior preferences. A change in conceptualisation would imply that the dementia patient develops or drops a value, or adds or deletes a new domain. Schölzel-Dorenbos et al4 have argued that comfort and safety matter most for dementia patients and that their intellectual capacities do not matter to them anymore. In other words, these intellectual capacities are not valuable to them anymore. It is true that many dementia patients do not express the importance of these capacities, and therefore these may seem to be unimportant to them, but it does not imply a change in their conceptualisation of what a good QoL is. We think that for the dementia patient the expression of pleasure and pain is not caused by a new conceptualisation of QoL, but shows the devastating effects of dementia that leave these patients no choice but to express themselves in these terms. The likeliness that dementia patients develop new values is rather unlikely, as also Jaworska acknowledges.21 A dementia patient might seem to act similar to a person who has consciously dropped his former values. However, it is highly questionable whether the dementia patient and the person who consciously dropped his values should be regarded the same. The change of mind should thus be understood as a new wish and not as a new preference or value.


It is a misconception to argue that Frank has undergone a response shift. Not all changes or lack of changes in perceived and/or self-reported QoL are driven by a response shift and we should refrain from using this concept to explain all changes in preferences/wishes.23 By definition, the term ‘response shift’ is only applicable when patients' internal standards, values or conceptualisation of QoL change. Concerns about the erroneous use of ‘response shift’ for other phenomena has been expressed previously,23 ,24 and conceptual clarity is required.

A dementia patient loses precisely the capacities that are necessary to change his mind deliberately and consciously. We should not argue that Frank does not want to be clever and in control anymore and that he does not mind being dependent, simply because he expresses different wishes from before. Frank's behaviour changes throughout the disease process, and he may seem to have adapted, but this is most likely caused by the loss of his ability to hold on to his prior preferences, not by a (sub)conscious choice to reset his standards, his values or the meaning he attaches to QoL. There are better explanations than a response shift for situations when dementia patients seem to ‘oppose’ their prior preferences. Changes in behaviour or a change of preference by dementia patients are caused by the disease itself, rather than by an adjustment to the disease.

The implications of our view are many, and merit extensive discussion. We will only highlight the important implications for advance directives. Even though not all wishes or values remain equally important once someone becomes incompetent, some prior preferences or values may carry extra importance, also or especially after becoming incompetent. An autonomous person, foreseeing that he might be unable to live according to his preferences in the future, could draw up an advance directive to make sure he will still be treated according to his current preferences and not according to his future wishes. Precisely because he has foreseen that he will not remember his values, he wants to make sure others will treat him according to these values in the future, also or especially when he has forgotten about them. It is unwarranted to argue that one's current well-being should always take precedence over all other values once a person is incapacitated.

Our paper shows that a response shift is the wrong model to explain the changes in preferences of dementia patients, but further analysis is required in order to assess when and under which conditions prior preferences take precedence over current wishes. In all situations in which current wishes seem to oppose prior preferences, further investigation is required. It should be investigated whether the change is genuine or whether the change is, for example, caused by temporary fear or confusion. Depending on these factors, and on the importance of the particular value and the importance of the issue at stake, a decision should be made whether prior preferences or current wishes should be considered most important.


Response shift is a wrong and misleading model for explaining why the wishes of dementia patients differ from their earlier preferences. Dementia interferes with the capacity for understanding, with the ability to remember and makes, especially in its advanced stages, self-evaluation impossible. Without self-evaluation a response shift is not possible. Furthermore, dementia patients lose capacities to hold on to prior preferences and may behave differently over time or may express conflicting wishes. This change is not caused by them dropping prior values and preferences, nor have they chosen to form new values or preferences. Therefore, a change in expressed wishes should not be followed blindly. Prior preferences may carry significance beyond the patient's competence, and explicit preferences that were directed to the future also matter when the patient has forgotten them. This is especially important for situations in which a person has expressed his preferences in an advance directive. These persons anticipated which preferences and values remain important to them in the case where they are no longer able to live accordingly. Advance directives of incompetent persons should not be followed blindly, but also cannot be simply put aside.



  • Contributors KRJ: substantial contributions to the conception and design of the work, drafting and revising it critically for important intellectual content; final approval of the version to be published; agrees to be accountable for all aspects of the work. MAGS and SvdV: substantial contributions to conceptual understanding, and interpretation of the paper, drafting and revising it critically for important intellectual content; final approval of the version to be published; agree to be accountable for all aspects of the work.

  • Funding ZonMw funded this work (40-41500-98-99002).

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Other content recommended for you