Emotional Memory Across the Lifespan: Positivity Effects, Reminiscence and Preservation in Dementia

Abstract

Does emotional memory decline with age as episodic memory does, or does it follow a distinct trajectory? The present review examines three bodies of evidence. First, socioemotional selectivity theory and the age-related positivity effect indicate that older adults preferentially attend to and remember positive material, a shift that a meta-analysis of 100 studies attributes to motivated, resource-dependent processing rather than neural decline. Second, research on reminiscence and life review shows that recollection in later life is functionally heterogeneous and that its benefits depend on which function is engaged; the reminiscence bump identifies early adulthood as the period from which emotionally salient memories are most densely retrieved. Third, studies of amnesia and Alzheimer’s disease demonstrate that the experience of an emotion can outlast declarative memory for the event that induced it. We conclude that emotional memory in ageing is selected, reconstructed and partly dissociable from event memory. For emotion preservation, this implies that what should be captured is not a neutral record of events but the affective trace the person has built around them, and that this trace remains ethically consequential when the record itself is no longer accessible.

Keywords: positivity effect; socioemotional selectivity; reminiscence; life review; reminiscence bump; Alzheimer’s disease

1. Introduction

The dominant framework for cognitive ageing describes monotonic decline in fluid abilities, episodic memory and processing speed. Emotional functioning does not fit this picture: surveys consistently report that negative affect decreases from early adulthood to at least the seventh decade while positive affect and life satisfaction remain stable, the so-called “paradox of well-being” in old age. Memory researchers have documented age-related changes in what is remembered, and clinical neuropsychology has shown that emotional experience may persist when the declarative record of its cause is gone.

These literatures have developed largely in isolation. The present review brings them together around one question: what is the status of emotional memory in the ageing individual, and which of its properties would a preservation effort need to respect? Section 2 reviews the positivity effect and its replication status; Section 3 examines reminiscence as function and intervention, and the temporal distribution of autobiographical memory; Section 4 considers the dissociation of emotion and memory in amnesia and Alzheimer’s disease; Section 5 translates the findings into requirements for preservation.

2. Motivated cognition in later life

2.1 Socioemotional selectivity theory

Carstensen, Isaacowitz and Charles (1999) proposed that the perception of time remaining in life, rather than chronological age as such, organises the goals people pursue. When time is perceived as open-ended, knowledge-related goals dominate: acquiring information, expanding social networks, preparing for a distant future. When time is perceived as limited, emotionally meaningful goals take precedence: present affective states, close relationships and the regulation of emotion. Because age is strongly correlated with perceived time horizon, older adults typically show the second profile.

The theory predicts that manipulating time horizon should shift goals independently of age. Younger adults facing life-limiting illness or political instability show the preference for familiar over novel social partners that characterises older adults; asking older adults to imagine an extended future reduces that preference (Carstensen et al., 1999). Age-related emotional change is thus adaptive reprioritisation rather than a by-product of loss.

2.2 The positivity effect

Mather and Carstensen (2005) extended the theory to cognition, arguing that the same motivational shift should bias attention and memory towards positive information. In the anchoring memory study, Charles, Mather and Carstensen (2003) presented young, middle-aged and older adults with positive, negative and neutral pictures under incidental encoding and later tested free recall and recognition. Recall of negative images decreased across age groups, whereas memory for positive images was relatively spared, so the ratio of positive to negative images remembered increased with age. Eye-tracking and dot-probe studies showed a parallel attentional bias.

Two features are theoretically decisive. The effect depends on cognitive control: under divided attention, or when older adults are instructed to attend to negative content, it is reduced or abolished. And older adults with better executive function show it more strongly, not less, which is difficult to reconcile with amygdala or frontal decline. The effect therefore appears to reflect goal-directed regulation rather than an inability to process negative material (Mather & Carstensen, 2005; Carstensen & DeLiema, 2018).

2.3 Replication and boundary conditions

The positivity effect has seen failed replications and disputes over its magnitude. Reed, Chan and Mikels (2014) meta-analysed 100 studies comparing younger and older adults on preferences for positive over negative information and reported a reliable overall age difference in the predicted direction, of small magnitude. Crucially, the effect was substantially larger in studies that left participants free to process stimuli as they wished than in studies imposing instructions, time limits or divided attention, in which it approached zero. This is what a motivational account predicts and a deficit account does not.

Two limitations remain. Most evidence is cross-sectional, so cohort effects cannot be excluded, and the effect concerns relative preference: with threat-related or highly arousing stimuli the bias often disappears. The established finding is that, when unconstrained, older adults allocate memory resources preferentially to positive material; the shift is modest and depends on available control resources.

3. Reminiscence, life review and the structure of autobiographical memory

3.1 From clinical observation to functional taxonomy

Butler (1963) reframed the tendency of older people to revisit their past, then widely regarded as a symptom of decline, as a normative and potentially adaptive process: the approach of death prompts a return to consciousness of past experiences, particularly unresolved conflicts, which are re-examined and, under favourable conditions, reintegrated. The outcome is not uniformly positive; life review may result in acceptance, but also in guilt, anxiety or depression when reintegration fails. The formulation was clinical and unquantified, but it set the agenda for five decades.

Webster (1993) supplied the psychometric structure. The Reminiscence Functions Scale, validated on a sample spanning late adolescence to old age, identified eight factorially distinct reasons for recalling the past: boredom reduction, death preparation, identity, problem solving, conversation, intimacy maintenance, bitterness revival and teaching or informing others. Subsequent work has repeatedly found that bitterness revival and boredom reduction predict poorer mental health, whereas identity, problem solving and the prosocial functions predict better adjustment.

Westerhof and Bohlmeijer (2014), reviewing fifty years of research, distinguished simple reminiscence (unstructured, conversational), life review (structured, evaluative, integrative) and life-review therapy (life review directed at psychopathology), and noted a shift from asking whether reminiscence is beneficial to asking which functions, for whom, and under what conditions.

3.2 Interventions: meta-analytic evidence

Pinquart and Forstmeier (2012) meta-analysed 128 controlled studies of reminiscence interventions. Moderate improvements were found for depressive symptoms and ego integrity (Hedges’ g in the region of 0.6), with smaller but significant effects on purpose in life, death preparation, mastery, well-being and social integration. Effects on depression were larger for life-review therapy than for simple reminiscence and in participants with clinically relevant depression at baseline, and were maintained but attenuated at follow-up. Study quality was variable and moderator analyses rest on few studies. Nonetheless, structured recollection of the emotional past produces measurable changes in affective state, and structure and integration, rather than recollection per se, carry the effect.

3.3 The reminiscence bump

When adults over about forty produce autobiographical memories, their distribution across the lifespan shows a marked over-representation of memories from roughly ages 10 to 30. Rubin, Rahhal and Poon (1998) showed that this reminiscence bump is not restricted to personal episodes: older adults also disproportionately date their favourite films, books and music to this period, name public events from it as most important, and perform best on knowledge of news and cultural items from it. Early adulthood thus appears to be a privileged period of encoding whose products are retained across life.

Explanations are not mutually exclusive: novelty produces distinctive encoding; identity formation makes these memories central to the self and frequently rehearsed; and culturally shared life scripts guide retrieval towards the period when important, largely positive, transitions are expected. Across explanations, the bump is dense in emotionally significant, self-relevant and predominantly positive events, and thus indicates where in a life the affectively central material is most likely to be found.

4. Emotion without memory: amnesia and Alzheimer’s disease

4.1 Sustained feeling after loss of the inducing event

Feinstein, Duff and Tranel (2010) tested whether an emotional state requires memory of its cause in order to persist. Five patients with severe amnesia following bilateral hippocampal damage, and matched healthy participants, watched film clips inducing sadness and, in a separate session, happiness, with emotional state rated at intervals over the following half hour. The patients retained little or no memory of the clips, several failing to recall that they had watched a film at all. Nonetheless their sadness persisted for twenty to thirty minutes, comparable to or exceeding that of controls, and its persistence was inversely related to the amount remembered; happiness also persisted, though less consistently. The experience of an emotion thus does not depend on the declarative record of the event.

Guzmán-Vélez, Feinstein and Tranel (2014) replicated the design in patients with probable Alzheimer’s disease and healthy older adults. The patients forgot the film content within minutes, yet both sadness and happiness remained elevated throughout the observation period, and poorer memory was again associated with more prolonged sadness. The authors framed the result as “feelings without memory” and drew the clinical implication directly: interactions a patient will not remember can nevertheless produce durable emotional consequences in either direction. The finding is consistent with a neuropathology in which medial temporal structures supporting episodic memory are affected early while circuits supporting affective response are comparatively preserved. Both studies rest on small samples and laboratory induction, and neither identifies the mechanism maintaining an unattributed emotion; the direction of the effect, however, has been replicated across two aetiologies.

4.2 Emotional enhancement of memory in Alzheimer’s disease

A distinct question is whether emotional content still enhances memory when memory itself is impaired, as it does in healthy adults through amygdala modulation of hippocampal consolidation. Kensinger, Anderson, Growdon and Corkin (2004) examined memory for emotional words in Alzheimer’s disease relative to healthy older adults. The enhancement observed in controls was attenuated in patients, and in some conditions absent, depending on stimulus type and retrieval measure. Emotional enhancement appears to require an interaction between amygdala and medial temporal systems: when the hippocampal component is compromised, the amygdala continues to generate affective responses at encoding but can no longer confer an advantage on consolidation. This reconciles the two sets of findings: in Alzheimer’s disease, emotion is experienced and sustained, but it no longer reliably rescues the memory of what caused it.

4.3 Autobiographical memory and its emotional residue

El Haj, Antoine, Nandrino and Kapogiannis (2015) reviewed autobiographical memory in Alzheimer’s disease and characterised a shift from episodic to semantic retrieval: patients increasingly produce general or schematic memories in place of specific, contextually detailed episodes, with a temporal gradient sparing remote over recent periods; memories from the reminiscence bump are among the most resilient. The same group has reported that self-selected music increases the specificity and emotional content of autobiographical retrieval in mild Alzheimer’s disease, such memories being predominantly involuntary and positive. What survives longest in the autobiographical record of a person with dementia thus appears to be a small set of remote, emotionally weighted, self-defining memories, most readily reached through affectively charged cues rather than direct questioning. Most of these studies are small and await replication.

5. Implications for emotion preservation

The evidence reviewed can be stated as constraints on any preservation programme.

First, the emotional record of an older adult is a selected record. The positivity effect and the functional analysis of reminiscence show that what is retained and rehearsed reflects current goals, particularly affect regulation. Treating this selection as distortion to be corrected would misdescribe the object. We propose as a requirement that preservation capture the person’s evaluative stance towards an event alongside the event itself, and record the circumstances of retrieval (spontaneous or cued, conversational or evaluative) as part of the data.

Second, emotional memory is partly dissociable from event memory, and the affective component is the more durable. This yields a testable proposition: material collected from an individual before cognitive decline, if it encodes emotional appraisal and not only factual content, should continue to elicit corresponding affective responses after episodic access to the underlying events is lost. The reminiscence bump specifies where such material is most likely to be found, and the music findings in Alzheimer’s disease suggest which cue modalities are most likely to reach it.

Third, eliciting emotional memory is not neutral. Structured, integrative recollection improves affective outcomes, whereas recollection oriented towards bitterness predicts poorer adjustment. Elicitation protocols used for preservation should therefore be designed and evaluated as interventions, with the function engaged as an independent variable and well-being as an outcome.

Fourth, the persistence of unattributed emotion in dementia imposes an ethical constraint that precedes any collection. A person who cannot remember an interaction still bears its emotional consequences. Preservation activities involving people with dementia must be assessed for their affective footprint on the participant, not only for what is obtained. Open questions include whether preserved affective material can be transmitted to another person in a form that reproduces rather than describes the emotion, and how consent for such transmission should be understood when the source can no longer evaluate it.

6. Conclusion

The three literatures reviewed here describe one system from different angles. Socioemotional selectivity accounts for why older adults reorganise memory around emotional value; the reminiscence literature describes how that reorganisation is enacted in recollection and to what effect; and the dementia studies show what remains when episodic memory fails. The picture is of an emotional memory that is motivated, functionally differentiated and more robust than the events it accompanies. Several central findings rest on small or cross-sectional samples. Integrating experimental, gerontological and neuropsychological methods within longitudinal designs is the most direct route to establishing which properties of emotional memory are invariant across the lifespan and which depend on the goals, resources and health of the remembering person.

References

Butler, R. N. (1963). The life review: An interpretation of reminiscence in the aged. Psychiatry, 26(1), 65–76.

Carstensen, L. L., & DeLiema, M. (2018). The positivity effect: A negativity bias in youth fades with age. Current Opinion in Behavioral Sciences, 19, 7–12.

Carstensen, L. L., Isaacowitz, D. M., & Charles, S. T. (1999). Taking time seriously: A theory of socioemotional selectivity. American Psychologist, 54(3), 165–181.

Charles, S. T., Mather, M., & Carstensen, L. L. (2003). Aging and emotional memory: The forgettable nature of negative images for older adults. Journal of Experimental Psychology: General, 132(2), 310–324.

El Haj, M., Antoine, P., Nandrino, J. L., & Kapogiannis, D. (2015). Autobiographical memory decline in Alzheimer’s disease, a theoretical and clinical overview. Ageing Research Reviews, 23, 183–192.

Feinstein, J. S., Duff, M. C., & Tranel, D. (2010). Sustained experience of emotion after loss of memory in patients with amnesia. Proceedings of the National Academy of Sciences, 107(17), 7674–7679.

Guzmán-Vélez, E., Feinstein, J. S., & Tranel, D. (2014). Feelings without memory in Alzheimer disease. Cognitive and Behavioral Neurology, 27(3), 117–129.

Kensinger, E. A., Anderson, A., Growdon, J. H., & Corkin, S. (2004). Effects of Alzheimer disease on memory for verbal emotional information. Neuropsychologia, 42(6), 791–800.

Mather, M., & Carstensen, L. L. (2005). Aging and motivated cognition: The positivity effect in attention and memory. Trends in Cognitive Sciences, 9(10), 496–502.

Pinquart, M., & Forstmeier, S. (2012). Effects of reminiscence interventions on psychosocial outcomes: A meta-analysis. Aging & Mental Health, 16(5), 541–558.

Reed, A. E., Chan, L., & Mikels, J. A. (2014). Meta-analysis of the age-related positivity effect: Age differences in preferences for positive over negative information. Psychology and Aging, 29(1), 1–15.

Rubin, D. C., Rahhal, T. A., & Poon, L. W. (1998). Things learned in early adulthood are remembered best. Memory & Cognition, 26(1), 3–19.

Webster, J. D. (1993). Construction and validation of the Reminiscence Functions Scale. Journal of Gerontology, 48(5), P256–P262.

Westerhof, G. J., & Bohlmeijer, E. T. (2014). Celebrating fifty years of research and applications in reminiscence and life review: State of the art and new directions. Journal of Aging Studies, 29, 107–114.