Narcissism Vulnerability as Risk Factor for Alzheimer´s Disease- A Prospective Study

Research Article

Austin J Clin Neurol 2015;2(7): 1057.

Narcissism Vulnerability as Risk Factor for Alzheimer´s Disease- A Prospective Study

Serrani D*

Department of Psychobiology, National University of Rosario, Argentina

*Corresponding author: Daniel Serrani, Department of Psychobiology, Faculty of Psychology, National University of Rosario, Berutti 250 bis, Rosario, Santa Fe (S2000QBE), Argentina

Received: April 06, 2015; Accepted: June 25, 2015; Published: June 30, 2015


Introduction: We examine whether broad factors and specific facets of pathological narcissistic personality are associated with increased risk of incident Alzheimer’s disease (AD) in a prospective cohort study. Participants (n=452) were monitored for up to 36 months from a baseline personality assessment with Pathological Narcissistic Inventory (NPI). The results showed that dementia developed in 159 participants. On average, participants in whom dementia developed were mostly female (p<;.02), slightly older (p<.04), had lower levels of education (p<.03), lower scores on cognitive test (p<.01), higher scores on narcissistic vulnerability scores (p< .01); but no differences in depression score. Adjusted hazard ratios (AHR) and 95% Confidence Interval (CI) for AD were greater for the following subscales of the PNI: Hiding Self [AHR (95% CI) =1.34 (1.18-1.39)], Devaluation [AHR (95% CI) =1.35 (0.65, 0.98)] and Narcissistic Vulnerability [AHR (95% CI) =1.43 (1.34-1.69)]. Also cognitive status and depression showed robust association with incident AD. The current study indicates that narcissistic vulnerability traits are associated with increased risk of AD, with significant hazard ratios.

Keywords: Narcissistic personality; Alzheimer’s disease; Risk of dementia;Depression; Cognitive status


Alzheimer’s disease and related dementias (ADRD) represents an increasing burden for patients and caregivers, including physical, emotional, and financial demands for care provision [1-3].

An estimated 5.4 million Americans of all ages had Alzheimer’s disease in 2012, one in eight Americans over age 65, ranging from approximately 53 new cases per 1,000 people aged 65 to 74, to 170 new cases per 1,000 people aged 75 to 84. Global prevalence of dementia is as high as 24 million, and is expected to quadruple by the year 2050. In the US alone, it accounts for an estimate direct health-care costs of $172 billion per year and $210 billion in unpaid care giving [4].

Even when etiological mechanisms underlying this illness remain unknown, both environmental and genetic factors have been proposed as putative causes, beginning years before symptoms became clinically evident. Risk factors for dementia, such as advancing age, family history, genetics, brain reserve, and lifestyle have been proposed [5].

Provided that only modest therapeutic effects have been achieved with currently available pharmacologic disease-modifying treatments for AD, attention has shifted toward prevention or delay of AD onset [6,7].

In a recent systematic review, association of multiple factors with ADRD was assessed, and several potentially modifiable conditions were identified, such as lifestyle, diabetes, hypertension, obesity, smoking, depression, cognitive engagement, physical activity and diet. Surprisingly, personality traits were not addressed [8] despite strong associations between personality and health [9].

Very few studies have evaluated the relationship between premorbid personality, life events and ADRD. For example, early childhood stressful situations could impair elder´s psychosocial adaptation favoring the emergence of ADRD [10].

Previous psychiatric history, including premorbid personality, may predispose subjects to ADRD and its corresponding psychiatric symptoms [11-13] concluding that before Alzheimer’s disease onset, patients were more passive, hostile and less spontaneous. Other authors also reported diminished initiative/growing apathy, relinquishment of hobbies and increased rigidity [14].

Similar results were obtained in studies comparing premorbid and present personality using the NEO-PI instrument [15]. Changes in the five personality domains could be summarized as higher neuroticism, lower extraversion and conscientiousness, and smaller reductions in openness and agreeableness.

In a long term prospective study, higher scores on vulnerability to stress, anxiety, and depression preceded onset of dementia up to 30 years. Those traits were associated with higher ADRD neuropathology at autopsy, and lower resilience to clinical dementia. On the other side, agreeableness, order and competence, as part of a resilient personality, were significantly associated with lower risk or delay of clinical dementia, even when post mortem AD neuropathology was present [16,17]. Those findings were replicated in an independent study [18].

In another research using data from the Religious Orders Study [19] conscientiousness or will and goal directedness were associated with a reduced risk of AD, and to a slower rate of cognitive decline.

Several other studies have acknowledged the association between mid-life personality and incident ADRD [20-24]. Finally, in recent meta-analysis [25] individuals with high neuroticism and low conscientiousness had a threefold increased risk of incident AD, together with self-discipline and depression. Some studies do not support those previous findings, however, and found no relationship between previous personality and risk of dementia [26].

Some authors, indeed, tend to stress the continuity between preand post-morbid personality profiles [27,28]. For those researchers, the personality traits of patients with dementia seemingly reflect adaptive coping models used before onset of dementia. Considering the aforementioned studies, research on personality disorder as a risk factor for dementia usually has been based on psychobiological models [29] or the Five Factor Taxonomy (FFT) of personality [30- 32], and examined relatively homogeneous groups [33].

Given the putative role of education and lifestyle as risk factors for ADRD [34], research should also focus in ethnically heterogeneous cohorts and tap on more clinically oriented personality clusters. In that sense, subjects with personality disorders and associated neurobiological vulnerabilities, such as Borderline, Narcissistic or Antisocial Personality Disorder [35,36] could be prone to develop neurodegenerative diseases such as dementia in old age.

However, only few case reports are available to support this hypothesis [37-39] without corresponding epidemiological data. With regard to Narcissistic Personality Disorder, the most consistent findings have been a positive correlation with negative emotionality, aggressiveness, psychoticism and extraversion; a negative correlation with agreeableness and conscientiousness36, and inconsistent correlation with introversion and low positive emotionality.

A substantial degree of variability in convergence between measures of narcissism grandiosity and vulnerability and the domains of the FFT still persist [40], suggesting that narcissistic personality as risk factor for dementia should be studied in its own, and not simply associating it with the FFT results [41]. As most of the evidence on risk factors for ADRD has been based on case-control studies, longitudinal studies may represent a better design to identify risk factors for development of ADRD of affected individuals. They avoid the need to obtain exposure data from proxy respondents as individuals affected with ADRD may be unable to provide reliable data.

In conclusion, the objective of this research was to investigate the role of previous pathological narcissistic personality disorder as a risk factor for ADRD in a Spanish-speaking population using a longitudinal study. Ultimately, we hope that these data could aid in two ways: first, developing clinically useful risk markers for predicting ADRD, which, in turn, would pave the way for dementia prevention strategies; second, studying the risk posed by personality could yield insights about the etiology of ADRD [42-44].

Materials and Methods

Sample size

Based on an age-standardized prevalence for ADRD of 4.4% [45- 47], a relative risk (RR) of 1.19 in relation to premorbid health status, an alpha risk of 5%, a power of 95% and an expected drop-out rate of 20% over 36 months, 452 patients were selected.

Study design

The research was based on a prospective cohort study of Spanishspeaking subjects with personality disorders who resided in the community, and received psychiatric treatment at a community mental health center. Inclusion criteria included age 65-75 years, availability of a proxy respondent, preserved ability to complete clinical and neuropsychological evaluations, sign informed consent, and exhibit normal cognitive performance and no significant morbidity. Exclusion criteria included severe visual or hearing impairment, dementia, idiopathic Parkinson’s disease, liver disease, alcoholism, known terminal illness, hospitalization for depression within the last year or having received electroconvulsive therapy within the prior decade, current treatment with cholinesterase inhibitors, anti-Parkinson medications, tricyclic antidepressants, antipsychotics or other medications with significant psychotropic or central cholinergic effects. Patients were enrolled and interviewed from March 2009 to March 2010. At baseline, patients underwent a clinical, neuropsychological and psychiatric examination, and a brain MRI. Patients were followed up over a period of 36 months, at 6 months intervals. Informed written consent was obtained from subject and caregiver at enrollment and before baseline assessment took place. The study protocol was reviewed and approved by the local institutional review board. All procedures were in accordance with the declaration of Helsinki. The sample was middle-class, most subjects were female (63 percent). Only 2 subjects with missing data, and 7 subjects who moved or declined to return for follow-up were excluded. After that, 452 subjects remained available for the study.


Cognitive impairment was assessed using the Modified Mini- Mental State Exam (normal values ≥78) [48]. The 3MSE is based on the Mini-Mental State Exam and offers a more graded scoring.

Depression was assessed using the Geriatric Depression Scale (Short Form) [49] which has shown good psychometric properties. A score of 0 to 5 is normal. A score greater than 5 suggests depression.

Diagnosis of dementia was made using the informant/subject structured interview based on the Clinical Dementia Rating Scale (CDR) (normal values ≤ 0.5) [50] and the Blessed Information Memory Concentration scale (normal values ≤ 4) [51]. This last test exhibits a high test–retest reliability (0.86), and correlate closely with the stages of Alzheimer’s disease [52].

Narcissistic Pathological Personality was assessed with the Pathological Narcissism Inventory (PNI) [53] which is a 52-item multidimensional self-report measure of pathological narcissism. Respondents are asked to use a 6-point scale ranging from 0 (not at all like me) to 5 (very much like me) to rate each item. It consists of seven subscales that measure different characteristics of pathological narcissism: Contingent Self-Esteem (CSE), Exploitativeness (EXP), Self-Sacrificing Self-Enhancement (SSSE), Hiding the Self (HS), Grandiose Fantasy (GF), Devaluing (DEV), and Entitlement Rage (ER). Because of the variability in scale length, mean scores are used instead of sums to allow for easy comparison across scales. Two higher order factors encompassing these 7 subscales, Narcissistic Grandiosity and Narcissistic Vulnerability are also scored.

Statistical analysis

Data were analyzed using Stata (Stata Statistical Software: Release 13. College Station, TX: StataCorp LP). Continuous variables were compared with use of either an independent-samples t-test or the Mann–Whitney U test, and categorical variables were compared with use of the Pearson chi-square test. The association between personality and risk of dementia was assessed using cox proportionalhazards regression analysis to estimate hazard ratios, with 95 percent confidence intervals. The time to an event was defined as time evolved from enrollment to the date when a diagnosis of dementia was issued. Censoring was based on study ending, time of death or follow-up refusal. All multivariate models included following covariates: age at enrollment, sex, educational level (high school vs. college-level education) and base-line scores of 3MSE and geriatric depression scale. Graphical (log-log plots) and analytical (Schoenfeld residuals) methods were used to test the proportional hazards assumption for each analysis; the assumption was supported in every case. Analyses also examined interactions between each of the PNI subscales and other covariates in the model. Finally, analyses explored curvilinear relationships between the PNI subscales and dementia risk, by including each factor together with a squared term for each factor.


During 1104 person-years of follow-up (median follow-up, 2.7 years), dementia developed in 159 participants (Alzheimer’s disease in 112 (70%), vascular dementia in 37 (23%), mixed dementia in 9 (1.5%), and other types of dementia in 1 (0.5%). By the end of the study period, no subjects had died and 7 participants had dropped out (follow-up, 2.7±1.1 years). On average, participants in whom dementia developed were mostly female (t=3.7, p<.02), slightly older (t=4.9, p<.04), had lower levels of education (t=5.2, p<.03), lower scores on cognitive test (t=4.9, p<.01), higher scores on narcissistic vulnerability scores (t=8.7, p< .01); but no differences in depression score (Table 1).