Personality and Alzheimer’s disease: An integrative review

Abstract

The objective of this article is to review and integrate interrelated areas of research on personality and Alzheimer’s disease (AD). Prospective studies indicate that individuals who score higher on conscientiousness (more responsible and self-disciplined) and lower on neuroticism (less anxious and vulnerable to stress) have a reduced risk of developing dementia, even in the presence of AD neuropathology. Personality is also related to measures of cognitive performance and cognitive decline, with effect sizes similar to those of other clinical, lifestyle, and behavioral risk factors. These associations are unlikely to be due to reverse causality: Long-term prospective data indicate that there are no changes in personality that are an early sign of the disease during the preclinical phase of AD. With the onset and progression of dementia, however, there are large changes in personality that are reported consistently by caregivers in retrospective studies and are consistent with the clinical criteria for the diagnosis of dementia. The review also discusses potential mechanisms of the observed associations and emphasizes the need for prospective studies to elucidate the interplay of personality traits with AD neuropathology (amyloid and tau biomarkers) in modulating the risk and timing of onset of clinical dementia. The article concludes with the implications of personality research for identifying those at greater risk of AD and the potential of personality-tailored interventions aimed at the prevention and treatment of AD.

Alzheimer’s disease (AD) is the most common cause of dementia and the fastest-growing leading cause of death in the United States. AD is a progressive neurodegenerative disease that often manifests first as mild cognitive impairment (MCI) and then dementia. It is characterized by loss of memory and other cognitive functions, and changes in behavior, mood, and personality. These losses interfere with a person’s daily life and pose a significant burden to families and the healthcare system. Besides AD, other types of dementia include Lewy body, frontotemporal, and vascular dementia.

This article reviews and integrates the literature on personality as a risk factor for dementia, personality change in people with dementia, and potential pathways that may explain the observed associations. We further point to research questions that need to be addressed to advance knowledge on personality and dementia and consider the implications for the diagnosis and treatment of AD. Consistent with the literature on personality and dementia, this review focuses on research based on the five-factor model of personality (FFM).

Personality and Risk of Dementia and Cognitive Decline

Dementia risk.

Other than rare cases caused by specific genetic mutations, AD is a complex, multifactorial disease. Age is the strongest risk factor for AD, with increasing incidence from 1–2% for individuals aged 60–65 to about 30% for those over the age of 85. The e4 variant of the APOE gene is another strong risk factor, increasing the risk by 8- to 12-fold in e4 homozygous individuals compared to non-carriers. Other prominent risk factors can be grouped broadly under cardiovascular (e.g., midlife obesity, midlife hypertension, and diabetes), behavioral (e.g., smoking and physical inactivity), and psychosocial (depression, social contact, and education) risk (Beydoun et al., 2014; Chuang et al., 2015; Whitmer, Sidney, Selby, Johnston, & Yaffe, 2005).

Personality traits have robust links with psychological, behavioral, and clinical risk factors for AD (Chapman, Roberts, & Duberstein, 2011; Hampson, 2012). For example, neuroticism is a strong risk factor for major depression (Kendler, Gatz, Gardner, & Pedersen, 2006) and conscientiousness is related to cigarette smoking, physical inactivity, and obesity (Sutin et al., 2016). Personality traits are also related to educational achievements, coping skills, and the quality and quantity of interpersonal relations, which in turn are associated with increased vulnerability to dementia (Beydoun et al., 2014; Crooks, Lubben, Petitti, Little, & Chiu, 2008).

These links have led to the hypothesis that personality traits are associated with the risk of developing AD and related dementias. A recent meta-analysis (Terracciano et al., 2014) summarized the results of five prospective studies that included up to 5,054 individuals (Duberstein et al., 2011; Terracciano et al., 2014; Wilson et al., 2006; Wilson et al., 2005; Wilson, Schneider, Arnold, Bienias, & Bennett, 2007). At baseline, individuals free of cognitive impairment completed a self-report version of the NEO personality questionnaire. These individuals were then followed over time to investigate whether personality traits predicted who developed clinical dementia. There was consistent evidence that individuals who scored higher on neuroticism or lower on conscientiousness had greater risk of incident AD. Although there was weak evidence for an association in the individual studies, the meta-analysis indicated that higher openness and agreeableness were associated with a slightly reduced risk of AD. Extraversion was unrelated to AD risk. The effects of neuroticism and conscientiousness were independent of each other and there was not an interaction between the two traits (Terracciano et al., 2014; Wilson et al., 2005). Two studies examined the facets of neuroticism and found that anxiety and vulnerability to stress were significant predictors of incident AD (Terracciano et al., 2014; Wilson, Begeny, Boyle, Schneider, & Bennett, 2011). Two studies that examined the facets of conscientiousness found that several of its facets predicted incident dementia, with low self-discipline and responsibility being the strongest predictors (Sutin, Stephan, & Terracciano, in press; Terracciano et al., 2014).

The effect of neuroticism and conscientiousness are not just consistent across studies, but are also of non-negligible magnitude. In the Religious Order Study (Wilson et al., 2007), for example, those scoring in the bottom 10% of conscientiousness had about double the risk of AD as compared to those in the top 10%. Similarly, in the Baltimore Longitudinal Study of Aging (Terracciano et al., 2014), individuals with scores in the top quartile of neuroticism or the lowest quartile of conscientiousness had a three-fold increased risk of incident AD. These effect sizes are similar or larger than those found for cardiovascular and behavioral risk factors, such as diabetes, midlife obesity, midlife hypertension, lower education, physical inactivity, and cigarette smoking (Barnes & Yaffe, 2011; Daviglus et al., 2011).

Cognitive performance, decline, and impairment.

Consistent with prospective studies that examined incident AD, personality traits are associated with cognitive decline and mild impairments that can culminate in dementia. A large study of 4,039 members of the Swedish Twin Registry found neuroticism, but not extraversion, to be associated with cognitive impairment status 25 years later (Crowe, Andel, Pedersen, Fratiglioni, & Gatz, 2006). In cognitively healthy older adults, personality traits have been associated with the level of cognitive performance and rate of cognitive decline (Caselli et al., 2016). For example, in the Health and Retirement Study (N=13,987), individuals who scored lower in neuroticism or higher in openness and conscientiousness performed better on a memory task (Luchetti, Terracciano, Stephan, & Sutin, 2015). A meta-analysis indicated that higher neuroticism and lower conscientiousness were associated with steeper cognitive decline over time. Although the association between personality and cognitive decline over short follow-up periods was small, the effects were stronger than the effect of cardiovascular conditions, physical inactivity, and history of smoking and psychological distress (Luchetti et al., 2015). The protective effects of high conscientiousness are particularly strong in predicting the rate of cognitive decline within the last three years before death compared to the pre-terminal decline (Wilson et al., 2015). Finally, personality is also associated with subjective cognition. Individuals who score higher in neuroticism or lower in conscientiousness report more cognitive complaints (Steinberg et al., 2013) and worse self-rated memory (Luchetti et al., 2015). Subjective cognitive complaints are associated with increased risk of incident AD (Geerlings, Jonker, Bouter, Adèr, & Schmand, 2014).

Personality Disorder.

In addition to normal personality traits, personality disorders have also been associated with dementia. Most of this evidence is from studies with retrospective assessments by family members. Obsessive–Compulsive Personality Disorder, for example, is associated with increased risk of AD (Dondu, Sevincoka, Akyol, & Tataroglu, 2015); other studies report associations with almost all personality disorders (e.g., Nicholas et al., 2010).

Pathways Linking Personality to Dementia

Mechanisms.

As mentioned above, personality traits are related to multiple lifestyles and health factors that increase risk of dementia, and these factors are potential behavioral and clinical mechanisms through which personality is associated with dementia risk. In particular, personality may contribute to risk of dementia by shaping over the lifetime an individual’s reactions to stress, health behaviors, and engagement in physical, cognitive, and social activities. The finding that personality is associated with inflammatory markers (Luchetti, Barkley, Stephan, Terracciano, & Sutin, 2014) and brain-derived neurotrophic factor (Terracciano et al., 2011) suggests potential biological pathways. Still other studies have tested to what extent the association between personality and incident AD or cognitive decline was accounted for by other risk factors such as education, APOE genotype, activity patterns, or vascular conditions. Although attenuated, the associations between personality and cognition-related outcomes generally remained significant when accounting for such factors (Luchetti et al., 2015; Sutin et al., in press; Terracciano, Stephan, Luchetti, Albanese, & Sutin, 2017b; Wilson et al., 2003; Wilson et al., 2007). Some studies control for measures of distress or depressive symptoms when examining neuroticism as a predictor of incident AD (Johansson et al., 2014), but these findings are difficult to interpret given the conceptual and measurement overlap between neuroticism and other measures of distress. Rather than confounding factors, the psychological, behavioral, and clinical variables are likely to be partial mediators, and thus mechanisms, of the association between personality and incident AD.

Moderators.

A few studies have tested whether personality interacts with other risk factors to modulate the effect of personality on cognitive outcomes. Neuroticism, for example, may moderate the effect of the APOE genotype on cognitive function and risk of dementia (Dar-Nimrod et al., 2012). Other studies, however, found different (Sapkota, Wiebe, Small, & Dixon, 2015) or no interactions (Terracciano et al., 2014). Personality may also interact with the environment to predict dementia risk (e.g., inactive or socially isolated lifestyle, Wang et al., 2009), but to date there is no robust evidence that the associations vary across demographic groups defined by age, sex, race, ethnicity, or education (Terracciano et al., 2017b).

Reverse causality.

Behavioral and personality changes are a common clinical sign of AD (McKhann et al., 2011), and changes in the pre-clinical phase could be responsible for the observed associations between personality and incident dementia. This reverse causality hypothesis is plausible because the neuropathological change in AD occurs years before dementia onset (Jack et al., 2013). The accumulation of amyloid-β (Aβ) and tau leads to neuronal dysfunction and cell death, which spreads through the brain and is manifested in the clinical symptoms. Cerebrospinal fluid (CSF) and position emission tomography (PET) amyloid imaging indicate that Aβ accumulation occurs about a decade or more before clinical manifestation of the disease (Buchhave et al., 2012; Resnick et al., 2015). While there is a delay in the onset of cognitive impairment, it is possible that personality change occurs earlier. For example, individuals who later developed AD were cognitively normal but may have some Aβ deposition at the time they completed the personality questionnaire. The cascading neurodegeneration may have an impact on their brain and potentially their personality before the onset of cognitive symptoms. Of note, many studies on personality and incident AD were based on cohorts of older adults (mean age > 70 years) and had relatively short follow-up periods (≤6 years). As such, those who developed dementia during the follow-up period were likely in the AD prodromal phase when the personality questionnaire was completed.

Although plausible, there is clear evidence against the reverse causality hypothesis. A recent long-term prospective study (n = 2046) examined repeated NEO personality assessments over a time span of up to 36 years (Terracciano, An, Sutin, Thambisetty, & Resnick, 2017a). It found that the personality trajectories of those who developed AD were not significantly different from those who remained cognitively normal. Even within the last few years before the onset of MCI or dementia, the study found no evidence of preclinical changes in personality that could be interpreted as an early sign of AD (but see, Balsis, Carpenter, & Storandt, 2005; Duchek, Balota, Storandt, & Larsen, 2007). Other evidence against the reverse causality hypothesis includes findings that personality is associated with cognitive performance and predicts cognitive decline in relatively young samples (e.g., mean age = 45 years in Hock et al., 2014).

Personality resilience in the presence of AD neuropathology.

As noted above, Aβ dysregulation occurs years before the onset of dementia, and 20–40% of clinically normal older adults have biomarker evidence of AD neuropathology (Jack et al., 2013; Sperling et al., 2011). There is also considerable variability in the time lag between the emergence of AD neuropathology (e.g., positive PET amyloid imaging) and the onset of clinical dementia. Consistent with biomarker evidence, about 30% of individuals without cognitive impairment before death are found to meet the neuropathological criteria of AD at autopsy (Balsis et al., 2005; Driscoll & Troncoso, 2011). Personality traits may moderate the emergence of clinical signs of dementia in individuals with AD pathophysiology. We tested this hypothesis by comparing the personality traits of those who developed dementia with those who did not despite being found at autopsy to meet criteria for AD neuropathology. Compared to those with clinical dementia, asymptomatic individuals (those with AD neuropathology but no clinical dementia) scored higher on conscientiousness and lower on neuroticism (Terracciano et al., 2013). These findings suggest that a resilient personality profile plays a significant role at the interface of neuropathological processes and the manifestation of clinical symptoms. In people with underlying AD neurodegeneration, personality traits may help postpone the onset of clinical signs. Research with in vivo biomarkers (CSF or imaging) and concurrent personality assessments in large prospective studies is needed to confirm this hypothesis. Such research would also help clarify the temporal progression of personality change that occurs with the onset of other clinical signs and the evolution of AD biomarkers.

The evidence is mixed on the link between personality traits and neuropathology at autopsy. Higher neuroticism has been associated with more advanced spread of neurofibrillary tangles in limbic and neocortical regions, as indexed by Braak staging, but not with levels of Aβ plaques (Terracciano et al., 2013); other studies have found no association with neuropathology at autopsy (Wilson et al., 2003; Wilson et al., 2007). Data from in vivo brain imaging, indicates that individuals with MCI who score higher in neuroticism or lower in conscientiousness have more severe white matter lesions, but not cerebrovascular lesions or medial temporal lobe atrophy (Duron et al., 2014). In a large community-based cohort of about 500 older adults, individuals who scored lower in conscientiousness had more white matter hyperintensities, white matter fractional anisotropy, and brain-tissue loss (Booth et al., 2014).

Personality Changes and Dementia

Since Alois Alzheimer’s first case, personality change has been observed in individuals with AD and related dementias (Hippius & Neundorfer, 2003). Current National Institute on Aging and Alzheimer’s Association diagnostic guidelines indicate that the diagnosis of dementia requires impairment in at least two of five domains; one of the five domains is personality change, such as impaired motivation, social withdrawal, and increased irritability (McKhann et al., 2011).

Retrospective studies.

The common clinical observations and dementia guidelines are supported by research based on standardized measures of personality (Robins Wahlin & Byrne, 2011). Family members (usually a spouse or adult child) are asked to rate the premorbid (before the onset of dementia) and current personality of the individual with dementia. Good rater agreement is found when multiple informants are used (Strauss, Pasupathi, & Chaterjee, 1993), and current ratings are reliable and sensitive to change when repeated over time (Strauss & Pasupathi, 1994). Most studies have shown remarkably consistent patterns in terms of direction and magnitude of change. As illustrated in Figure 1, which is based on data from a meta-analysis of 9 studies (Robins Wahlin & Byrne, 2011), the differences between the premorbid and current ratings point to large changes in neuroticism (about 1.4 SD increase), extraversion (about 1.3 SD decrease), and conscientiousness (about 2.4 SD decrease). Smaller declines are also observed for openness and agreeableness (<0.5 SD). The differences in neuroticism, extraversion, and conscientiousness are large in absolute terms and are more than 10-fold larger than the changes typically observed with normal aging (Terracciano, McCrae, Brant, & Costa, 2005). It is worth noting that these studies compare premorbid to current personality. As such, the observed changes may have occurred after the onset of dementia, especially because prospective research has found no changes in the preclinical phase. Similar patterns of personality change have been observed in individuals with MCI (Donati et al., 2013), but the magnitude was considerably smaller (about 0.25 SD). Given the evidence to date, the changes in personality may start with MCI and become larger with the onset and progression of dementia.

Figure 1.

Personality change in individuals with dementia from 9 studies summarized by Robins Wahlin and Byrne (2011).

By the time of diagnosis, observer ratings of personality can discriminate healthy aging from early-stage AD beyond the performance on standard neuropsychological testing (Duchek et al., 2007). Personality change may also help differentiate AD from other types of dementia, such as frontotemporal dementia and Lewy Body dementia, to improve differential diagnosis. But the small studies that compared personality change across different types of dementia have found mixed results (Lykou et al., 2013; Torrente et al., 2014).

Prospective studies.

Compared to the dozen studies that examined personality change with a retrospective study design, fewer studies have used a prospective design to investigate personality change. As described above, a long-term longitudinal study has found no evidence of significant personality change in the preclinical phase of AD (Terracciano et al., 2017a). Similarly, a German longitudinal study (Kuzma, Sattler, Toro, Schonknecht, & Schroder, 2011) obtained self-reported personality ratings of 222 older adults three times over 12 years. At baseline, the participants were not cognitively impaired, but 66 were found to have developed MCI by the third follow-up. Contrary to expectation and a retrospective study (Donati et al., 2013), the group that developed MCI declined slightly on neuroticism, showed no change on extraversion (as compared to controls), and no change on conscientiousness (in the article, mean conscientiousness at t3 was mistyped as 2.3 instead of 3; Personal communication October 2015). One prospective study has found change in self-reported neuroticism, but not extraversion, in 86 octogenarians who received a diagnosis of dementia (Yoneda, Rush, Berg, Johansson, & Piccinin, 2016). Given that this latter study included the assessment at the time of the diagnosis, the increase in neuroticism is likely to have occurred at the time of the diagnosis. One prospective study over four years found that extraversion declined as the Clinical Dementia Rating (CDR) progressed from 0.5 (very mild) to ≥2 (moderate to severe disease stage) (Sollberger et al., 2011). Taken together, evidence from studies with a prospective design and the use of self-report data finds no evidence of change in the preclinical phase of the disease, while changes become evident at the time of the onset of dementia.

Self-reports vs. observer ratings.

As reviewed above, the research on personality change in individuals with dementia has relied on both observer ratings and self-reports methods. In non-clinical samples, there is generally a moderate agreement between self- and observer ratings (Funder, 1995). However, there is a discrepancy between the personality-as-a-risk-factor literature that is based on self-reports (Terracciano et al., 2014) and the personality-change literature that relies on observer ratings (Robins Wahlin & Byrne, 2011). The former indicates that people who later develop AD tend to score themselves as higher on neuroticism and lower on conscientiousness before the onset of dementia. By contrast, informants rate the premorbid neuroticism and conscientiousness in the normal range (Figure 1), which suggests that informants provide overly positive retrospective judgments. A contrast has also been found in studies that directly compared observer ratings with self-reports by individuals with mild dementia. For example, Richman (1989) found that nursing-home residents with mild dementia rated themselves as less neurotic and more extraverted, open, agreeable, and conscientious compared to how their family members rated them. Similar findings were obtained in at least two other studies, and in each case the difference between the groups with dementia and controls was smaller when relying on self-reports as compared to observer ratings (Duchek et al., 2007; Pocnet, Rossier, Antonietti, & von Gunten, 2011).

Although self-reports in individuals with advanced dementia are difficult to obtain and of dubious reliability, the ratings of individuals with early stage dementia cannot be dismissed as unreliable. High retest correlations (range of r_tt=0.76 for agreeableness to r_tt=0.95 for neuroticism) over one to two weeks support the reliability of self-reports in individuals with mild dementia (Richman, 1989). Internal consistency also remains relatively high among people with dementia (Terracciano, Stephan, Luchetti, & Sutin, 2017c). The discrepancy between self- and observer-report methods has been generally interpreted as impaired self-awareness (anosognosia) of ongoing personality change in individuals with MCI or dementia (Donati et al., 2013; Pocnet et al., 2011). For example, Rankin et al (2005) argued that individuals with dementia (especially frontotemporal dementia) may lose the ability to update their self-image. The affected persons may respond to personality questionnaires by retrieving information relevant to their former self more than reflecting on their current personality. Of note, if changes in personality are not registered in self-reports, the association between self-report personality and incident dementia cannot be due to changes in the traits as would be predicted by the reverse causality hypothesis.

A number of other motives and biases may contribute to the discrepancies between self- and informant-report methods. Although reliable across time and raters (Strauss & Pasupathi, 1994; Strauss et al., 1993), systematic biases cannot be ruled out with the observer-rating method, such as recall biases, hindsight bias, idealization of the person before the disease, and contrast effects when comparing premorbid to current personality. Current ratings are potentially influenced by the diagnosis or the labeling of dementia. When uncertain or lacking information, knowledge about the disease or stereotypes are likely to influence observer ratings (Hoerger et al., 2011). Long-term prospective studies with repeated assessments of both self- and observer-ratings are needed to tease apart these possibilities and determine at what stage of the disease process the two methods of personality assessment start to diverge. Such research would be most informative if the changes in personality can be mapped against changes in the underlying neuropathology (e.g., CSF or PET biomarkers) and other clinical signs.

Rank-order stability.

In addition to mean-level trajectories, the rank order stability of personality provides information on whether individual differences are maintained with the progression of the neurodegenerative disease. A prospective study that examined self-report personality found that the stability of all five traits to be substantially lower in the group with dementia compared to unimpaired old adults: the 4-year stability coefficient averaged across the five traits was .43 in the dementia group and about .70 in older adults with no dementia (Terracciano, Stephan, Luchetti, & Sutin, 2017c). In retrospective studies with observer-ratings, a few studies reported correlations ranging from .3 to .8 between the premorbid and current personality traits of individuals with mild to moderate dementia (see Terracciano et al., 2017c). These coefficients suggest that while personality becomes less stable, individuals with dementia retain some of their personality characteristics, at least in the mild to moderate stages of dementia. The persistence of some aspects of personality is of relevance for dementia care and underscores the continuity of personhood beyond the onset of dementia; personality can provide a framework to understand the behaviors, preferences, and values of the person with dementia.

Implications of Personality for diagnosis and interventions

Diagnostic utility.

Early diagnosis is crucial for the care of patients (HHS, 2012). There is also growing awareness that interventions need to occur before the disease progresses to advanced stages. It is therefore essential to identify AD cases at the preclinical stage of the disease, when disease-modifying interventions may be most effective (Sperling et al., 2011). Research on early markers of AD has focused on neuropathology, but Aβ and other biomarkers are only moderately correlated with clinical dementia (Holmes et al., 2008). In combination with biomarkers, personality traits may help identify individuals at greater risk of developing clinical dementia (Terracciano et al., 2013) and aid in the early diagnosis of AD (Duchek et al., 2007; Smith-Gamble et al., 2002). With the disease progression, personality provide a framework to interpret behavioral and psychological symptoms (Sutin, Stephan, Luchetti, Terracciano, in press).

Clinical trials.

High neuroticism and low conscientiousness are the two traits most associated with cognitive decline and risk of dementia. Yet, clinical trials likely miss individuals with these traits because they are less likely to volunteer for research studies (Lönnqvist et al., 2007). It is thus critical to consider personality traits in selection criteria for clinical trials, to select the individuals with the most vulnerable personality profile who are in the greatest need of effective treatments. As suggested by others (Duberstein et al., 2011), tailored incentives could improve recruitment of the population with personality-related vulnerability.

Intervention-driven personality change.

Estimates of population attributable risk suggest that neuroticism and conscientiousness may each account for about 10% of AD cases (Terracciano et al., 2014). Personality traits could thus be a promising target of interventions aimed at preventing or reducing the burden of AD. Interventions aimed at changing maladaptive aspects of personality are particularly enticing because these traits are thought to be distal causes of the disease and are linked to other risk factors and life outcomes. In addition to potential direct benefits, reducing neuroticism and increasing conscientiousness could have the added benefit of potentially reducing other risk factors, such as physical inactivity, social isolation, cigarette smoking, midlife obesity and other cardiovascular risk factors. Despite the potential direct and indirect benefits, relatively little is known about the effects of interventions on personality.

Perhaps the best evidence for intervention-induced personality change comes from pharmacological and psychoeducational treatments. For example, for individuals with major depression, antidepressants reduce the level of neuroticism (Costa, Bagby, Herbst, & McCrae, 2005; Tang et al., 2009). However, the side effects of psychotropic medications (e.g., benzodiazepines) in older adults can be particularly severe (Campanelli, 2012) and may aggravate more than reduce the risk of dementia (de Gage et al., 2012; Rosenberg et al., 2012). Besides pharmacological interventions, there is a vast literature on the benefits of cognitive behavioral treatments for anxiety and depression, which are likely to have an impact on neuroticism and other personality traits (see also Roberts et al, this issue). In the area of addiction, an outpatient drug-rehabilitation 6-week program study found significant changes (d ~ 0.3) in neuroticism, agreeableness, and conscientiousness that were maintained over a 15-month follow up (Piedmont, 2001). There is also increasing interest in cognitive training for personality change, but a recent study found that an intensive cognitive training for memory and perceptual speed had no positive impact on personality (Sander, Schmiedek, Brose, Wagner, & Specht, 2017). This finding might not be surprising given that cognitive training generally has limited transfer to other domains beyond the trained tasks. Interventions aimed at increasing physical activity (e.g., yoga, dancing), may have potential benefits for personality in addition to direct benefits for physical and cognitive health. In an observational study, older adults who were more physically active had a more favorable pattern of personality change (Stephan, Sutin, & Terracciano, 2014). However, it remains to be tested whether treatments designed to improve cognition, fitness, or other domains have broader and lasting benefits on personality.

Changing personality.

Although informative, pharmacological and psychoeducational interventions were done to treat depression, addiction, or to slow cognitive decline. In contrast, a recent study was designed specifically to change personality (Hudson & Fraley, 2015). In the first 16-week experiment, however, the intervention back-fired: those randomly assigned to a goal-setting intervention increased in neuroticism and decreased in agreeableness and conscientiousness. A second intervention based on more specific and concrete “change plans” was associated with better outcomes for neuroticism and conscientiousness but not for agreeableness (it decreased). Despite the use of college students, the study had high attrition (> 72%), which suggests that even among students there are significant barriers to consider, especially for those low on conscientiousness. The mixed results should not discourage further research. The bottom-up approach to changing behaviors, and, over time personality traits, has been advocated by others (Magidson, Roberts, Collado-Rodriguez, & Lejuez, 2014) and deserves further scrutiny. Such interventions are consistent with the individual’s desire to change (Hudson & Fraley, 2015) and the parental, school, workplace, friend, religious, institutional, and social pressures to change problematic behaviors and traits. Interventions can thus rely on tools that have been developed to address specific problems (e.g., smoking cessation), and test which tools might have an impact on personality over time. Furthermore, future research should examine whether observed effects in student samples can be replicated in adult samples, corroborated by observer ratings, and maintained over time. Interventions should be informed by evidence that personality is more malleable in children and adolescents, and it becomes increasingly stable in younger and especially in middle and older ages. Even after decades, stability coefficients are about r_tt=0.7 among adults older than 30 years (Terracciano, Costa, & McCrae, 2006). Such evidence suggests that interventions early in life might be more efficacious. Perhaps the most important and challenging question is whether experimentally induced alterations in self-reported personality traits have the desired effect on behavior change (e.g., more physical activity, quitting smoking) and health outcomes (English & Carstensen, 2014).

Personality-tailored interventions.

In the era of precision medicine (leveraging genetic and other information to enhance personalized interventions) and person-centered care (Edvardsson, Winblad, & Sandman, 2008), considering personality traits may be crucial for many aspects of dementia care. Most interventions tend to follow a “one-size-fits-all” approach that ignores individual differences. There is growing interest in models of care that recognize and respect the individual’s unique preferences, values, interests, and needs. Recognition of personality differences and other basic psychological dispositions is likely to shift the focus from the disease to the person, and to produce more effective interventions. For example, an intervention for risk reduction of alcohol misuse has shown that personality-targeted prevention programs are feasible and effective (Conrod et al., 2013). Similarly, personality-targeted interventions were effective in reducing depressive, anxiety, and conduct symptoms in high-risk youth (O’Leary-Barrett et al., 2013). Whether such personality-targeted intervention are more cost-effective than standard approaches remains to be determined. In dementia care, interventions that fit activities to individual preferences have been found to be more effective in reducing agitation and other behavioral symptoms (Gitlin et al., 2008; Kolanowski, Litaker, Buettner, Moeller, & Costa, 2011). In evaluating tailored interventions, more research is needed to examine whether the five broad factors or specific facets provide the most useful information for treatment. It is also important to include personality measures in large trials to test personality traits as predictors of treatment adherence and heterogeneity of treatment effects. This would provide an evidence-base on which interventions fit best given the psychological disposition of the person. Such knowledge would empower clinicians to provide care that is more person-centered.

Conclusion.

We reviewed evidence that conscientiousness and other personality traits may reduce risk of clinical dementia. We discussed several potential mechanisms for this protective role of personality, including indirect pathways by reducing other risk factors (e.g., physical inactivity, depression, etc.) and direct pathways by favoring brain health (Booth et al., 2014; Terracciano et al., 2011). In particular, we reviewed evidence that in the presence of AD neuropathology, personality may forestall the manifestation of clinical dementia. And contrary to the reverse causality hypothesis, long-term prospective data indicate that there are no preclinical changes in personality that may explain the association. We integrated these findings with evidence that dementia has an impact on personality, which constitutes a core clinical sign that emerges with the onset of dementia. We also point to the need for more prospective studies that include AD biomarkers, particularly to examine how personality modulates timing of disease onset. In conclusion, personality traits are a promising tool for the diagnosis, prevention, and treatment of dementia.