Caregiver Neuroticism Amplifies the Effect of Cognitive Impairment Severity on Caregiver Burden Among Dementia Caregivers
Article information
Abstract
Objective
Caregiver burden shows substantial variability across caregiving contexts. This study examined whether caregiver personality traits, particularly neuroticism, moderate the associations between care-recipient factors such as cognitive impairment severity and neuropsychiatric symptoms, and caregiving-related outcomes including overall burden and symptom-related distress.
Methods
Cross-sectional data from 214 care-recipient and caregiver dyads attending a geriatric psychiatry clinic were analyzed. Cognitive impairment severity was assessed using the Clinical Dementia Rating (CDR), and neuropsychiatric symptoms were measured with the Neuropsychiatric Inventory (NPI). Caregiving-related outcomes were evaluated using the Zarit Burden Interview (ZBI) and the NPI distress (NPI-D). Caregiver personality traits were assessed with the Big Five Inventory–Korean version. Linear regression models tested associations between care-recipient factors and caregiving-related outcomes, and multiple linear regression models examined moderation by caregiver personality using interaction terms, adjusting for caregiver age, sex, and cohabitation status.
Results
Caregiver neuroticism significantly moderated the associations between cognitive impairment severity and both overall caregiving burden and symptom-related distress by amplifying these associations (CDR×neuroticism interactions: ZBI, B [standard error, SE]=7.948 [3.387], p=0.020; NPI-D, B [SE]=4.498 [1.383], p=0.001).
Conclusion
Caregiver neuroticism appears to be a key modifier of how cognitive impairment severity translates into overall caregiving burden and symptom-related distress. Incorporating caregiver personality assessment may improve risk stratification and inform more personalized support approaches for caregivers of individuals with cognitive impairment.
INTRODUCTION
As the global population ages, the escalating prevalence of dementia poses a major public health challenge. Given the chronic and progressive nature of dementia, informal caregivers, primarily family members, play an essential role in supporting patients throughout the disease trajectory [1,2].
As reliance on informal caregiving for people with dementia intensifies, caregivers often experience considerable physical, emotional, and financial strain [3-5], collectively known as caregiver burden. This burden impacts caregivers’ health across physical and psychological domains, manifesting in adverse outcomes ranging from psychosocial issues like social isolation and depression [6,7] to elevated risks for cardiovascular disease, cognitive decline, and even dementia [8-12]. Alarmingly, caregiver burden may exacerbate neuropsychiatric symptoms in care-recipients, which in turn may contribute to broader clinical deterioration, including accelerated cognitive decline, thereby creating a mutually reinforcing cycle [13].
Building on this body of evidence, prior research has sought to identify factors that exacerbate or alleviate caregiver burden [14-17]. Clarifying the components of caregiver burden is therefore critical for delineating its distinct dimensions and underlying mechanisms.
Caregiver burden can be understood as encompassing two interrelated but distinct domains. The first involves the overall strain that arises from the emotional, physical, and financial demands of providing long-term care, which has long been recognized as a central aspect of the caregiving experience [17,18]. The second reflects the specific distress experienced in response to the behavioral and psychological changes commonly observed in care-recipients, representing the emotional toll of managing such neuropsychiatric manifestations [15,19,20]. Although these domains often exert cross-domain effects, distinguishing the general demands of care from symptom-specific distress is essential for the overarching direction of this study. By separating these elements, we can capture the interwoven patterns linking care-recipient factors with different facets of caregiver burden. This multidimensional approach ultimately allows us to determine whether caregiver personality traits moderate these distinct pathways differently.
Despite the established influence of care-recipient factors [10,16,21], there remains a gap in our understanding of how caregiver factors shape the experience of caregiver burden. A few studies examined caregiver personality as an independent predictor [22,23]. However, no prior study, to our knowledge, has examined whether caregiver personality acts as a moderator that conditions caregivers’ responses to clinical stressors, such as cognitive impairment and neuropsychiatric symptoms of the care-recipient. To address this gap, the present study investigated whether caregiver personality moderates the associations between care-recipient factors and caregiver burden, with a particular focus on neuroticism. This focus was informed by prior studies consistently identifying neuroticism as a primary vulnerability factor directly increasing caregiver burden [22,23]. We hypothesized that the associations between care-recipient factors and caregiver burden would be stronger among caregivers higher in neuroticism, with the goal of contributing to the development of more tailored support strategies.
METHODS
Study design and participants
This was a cross-sectional study utilizing data collected from the geriatric psychiatry clinic at Chungnam National University Hospital from August 2022 to August 2024. The study sample comprised 214 care-recipient–caregiver dyads. Eligible dyads included care-recipients who completed the CERAD-K (Consortium to Establish a Registry for Alzheimer’s Disease in Korea) clinical assessment and had a clearly identified primary informal caregiver. The primary informal caregiver was defined as the individual who self-identified as bearing the main responsibility for the patient’s daily care and who spent the most time with the patient. For caregivers who completed multiple surveys during the study period, only the most recent and complete entry was included to ensure temporal consistency between care-recipient and caregiver assessments. To ensure temporal alignment between assessments, only dyads in which the interval between the care-recipient’s clinical evaluation and the caregiver’s questionnaire completion was less than 1 year were included in the study dataset. In addition, dyads with missing data on core clinical or psychological measures, and in which the care-recipient had a global Clinical Dementia Rating (CDR) score of 0 were excluded. All participants provided written informed consent. The study protocol was approved by the Institutional Review Board of Chungnam National University Hospital (Reference Number: 2020-05-002), and all procedures adhered to the principles of the Declaration of Helsinki.
Care-recipient factors
We assessed the severity of cognitive impairment and neuropsychiatric symptoms as care-recipient factors.
Overall cognitive impairment severity was assessed using the CDR [24], a clinician-rated tool that evaluates functional performance across six domains including memory, orientation, judgment and problem solving, community affairs, home and hobbies, and personal care. Each domain is rated on a 5-point scale (0=none, 0.5=very mild, 1=mild, 2=moderate, 3=severe). The global score is calculated according to the established CDR scoring rules, which prioritize memory as the primary domain and incorporate patterns observed across secondary domains [24].
Neuropsychiatric symptoms were evaluated using the Korean version of the Neuropsychiatric Inventory (NPI), which captures both the frequency and severity of 12 neuropsychiatric domains [20,25]. Following the standardized NPI protocol, caregivers were asked about the presence of neuropsychiatric symptoms that had occurred within the past 4 weeks. For each symptom endorsed, frequency (rated from 1 to 4) and severity (rated from 1 to 3) were subsequently recorded. The domain score was calculated as the product of frequency and severity, yielding a range of 0 to 12. The NPI total score was computed by summing all domain scores and ranges from 0 to 144, with higher scores indicating more severe neuropsychiatric symptoms.
Caregiving-related outcomes
For a more precise analysis, the broadly used construct of caregiver burden was conceptualized in this study as caregivingrelated outcomes, encompassing multiple dimensions of caregivers’ experiences. These outcomes were operationalized into two standardized measures: overall caregiving burden, reflecting the global perceived strain associated with functional and cognitive care demands, and caregiver distress, representing affective burden elicited by care-recipient’s neuropsychiatric symptoms.
Overall caregiving burden was assessed using the Zarit Burden Interview (ZBI), a 22-item self-report questionnaire measuring the caregiver’s perceived burden associated with the functional and behavioral impairments of the care-recipient [26]. Each item is rated on a 5-point Likert scale (0 to 4), yielding a total score ranging from 0 to 88. Higher scores reflect greater caregiving burden. The Korean version of the ZBI has been validated and shows excellent internal consistency (Cronbach’s α >0.90) in studies involving caregivers of people with cognitive impairment.
The NPI distress (NPI-D) was used to quantify caregiver distress, capturing the emotional burden specifically associated with each neuropsychiatric symptom reported in the NPI [27]. After rating each symptom domain, caregivers were asked to rate their own emotional or psychological distress on a 6-point scale ranging from 0 (not at all distressing) to 5 (very severely distressing). The total NPI-D score is the sum of these ratings across all domains, yielding a range from 0 to 60, with higher scores indicating greater caregiver distress relate]d to neuropsychiatric symptoms [27].
Caregiver personality traits
Caregiver personality traits were assessed using the Big Five Inventory–Korean version (BFI-K), a validated 44-item selfreport questionnaire measuring five personality dimensions: extraversion, agreeableness, conscientiousness, neuroticism, and openness to experience [28,29]. Each item was rated on a 5-point Likert scale (1=strongly disagree to 5=strongly agree), and mean subscale scores were calculated for each trait. The BFIK was selected for its demonstrated reliability in Korean clinical and community settings [29].
Statistical analysis
Descriptive statistics were computed to summarize caregiver and care-recipient characteristics. Continuous variables are presented as means and standard deviations (SD), and categorical variables as frequencies and percentages. To examine the associations between care-recipient factors and caregiving-related outcomes, and to investigate the moderating role of caregiver personality, we conducted linear regression analyses to assess the independent associations of care-recipient factors (global CDR and NPI total) with caregiving-related outcomes (ZBI and NPI-D).
Next, to examine whether caregiver neuroticism moderated the associations between care-recipient factors and caregiving-related outcomes, multiple linear regression models were constructed for each outcome variable (ZBI and NPI-D).
Interaction terms were created for global CDR×neuroticism and NPI total×neuroticism, while controlling for age and sex of caregiver and cohabitation status. When significant interaction effects were identified, simple slope analyses were subsequently conducted to further explore the form of the interactions. For each significant moderator, the associations between the care-recipient factors and caregiving-related outcomes were probed at one SD above and below the mean of the personality trait.
To explore potential heterogeneity in the moderating effects of caregiver neuroticism across different caregiving contexts, subgroup analyses were conducted in a stratified manner according to caregiver sex, relationship to the care-recipient, and cohabitation status. Because these subgroup analyses were stratified by caregiver characteristics, the stratification variable was not included as a covariate to avoid redundancy and collinearity; instead, subgroup models were adjusted for carerecipient age and sex.
Analyses for the remaining four personality domains (extraversion, agreeableness, conscientiousness, and openness) were conducted in an exploratory manner to provide a comprehensive view of personality effects.
All analyses were conducted using Python (version 3.12.5; Python Software Foundation) and the statsmodels package (version 0.14.4; Statsmodels Development Team). Statistical significance was set at p<0.05 (two-tailed).
RESULTS
Participant characteristics
Demographic and clinical variables of both caregivers and care-recipients are summarized in Table 1. The mean age of caregivers was 61.0 years, and a majority were female (62.1%). In terms of their relationship to the care-recipient, spouses constituted the largest group (42.5%), followed by daughters (29.4%) and sons (18.7%). A majority of caregivers (58.4%) were co-resident with the care-recipient. Their mean score on the ZBI was 31.5, indicating a mild to moderate level of caregiving burden [26].
Participant demographics (N=214 dyads)
Associations between care-recipient factor and caregiving-related outcomes
We first examined whether cognitive impairment severity and neuropsychiatric symptoms were independently associated with ZBI and NPI-D. Higher global CDR was significantly associated with higher ZBI (B [standard error, SE]=14.475 [1.992], p<0.001) and NPI-D (B [SE]=5.424 [0.786], p<0.001). Higher NPI total scores were also linked to increased ZBI (B [SE]=0.917 [0.066], p<0.001) and NPI-D (B [SE]=0.511 [0.007], p<0.001) (Table 2 and Figure 1A).
Associations between care-recipient factors and caregiving-related outcomes
Conceptual models examining the association between care-recipient factors and caregiving-related outcomes. A: Linear regression model. This model illustrates the direct associations between care-recipient factors, including global CDR and NPI total, and caregiving-related outcomes, including ZBI and NPI-D. B: Moderation model. This model illustrates the moderating role of caregiver neuroticism in the relationship between care-recipient factors and caregiving-related outcomes. Specifically, global CDR and NPI total were examined for their associations with ZBI and NPI-D, while caregiver neuroticism was included as a moderator. The moderating effect of neuroticism was tested using multiple linear regression models including interaction terms between global CDR and caregiver neuroticism, and between NPI total and caregiver neuroticism. CDR, Clinical Dementia Rating; NPI, Neuropsychiatric Inventory; ZBI, Zarit Burden Interview; NPI-D, NPI distress.
Moderating effects of neuroticism on the association between global CDR and caregiving-related outcomes
We investigated whether neuroticism moderated the impact of care-recipient’s global CDR on ZBI and NPI-D (Table 3, Figures 1B and 2).
Moderating effects of caregiver neuroticism on the associations between care-recipient factors and caregiving-related outcomes
Moderating effect of neuroticism on the association between (A) Global CDR and ZBI and (B) Global CDR and NPI-D. Simple slope analyses are illustrated for high neuroticism (red line; mean +1SD) and low neuroticism (blue line; mean -1SD). CDR, Clinical Dementia Rating; ZBI, Zarit Burden Interview; NPI-D, Neuropsychiatric Inventory distress; SE, standard error.
For ZBI, neuroticism significantly moderated the association (B [SE]=7.948 [3.387], p=0.020). Subsequent simple slope analysis revealed that the effect of global CDR on ZBI was stronger at high neuroticism (mean +1SD; B [SE]=17.474 [2.591], p<0.001) compared to low neuroticism (mean -1SD; B [SE]=9.232 [2.784], p=0.011). This interaction indicates that for caregivers with higher neuroticism, the increase in ZBI associated with rising global CDR was significantly steeper than for those with lower neuroticism.
For NPI-D, neuroticism (B [SE]=4.498 [1.383], p=0.001) significantly moderated the association. Subsequent simple slope analysis revealed that the effect of global CDR on NPID was stronger at high neuroticism (mean +1SD; B [SE]=7.444 [1.037], p<0.001) compared to low neuroticism (mean -1SD; B [SE]=2.529 [1.112], p=0.024).
Moderating effects of neuroticism on the relationship between neuropsychiatric symptoms and caregivingrelated outcomes
We examined whether caregiver neuroticism moderated the impact of NPI total on ZBI and NPI-D (Table 3 and Figure 1B). Neuroticism did not significantly moderate the relationship between NPI total and either outcome.
Exploratory subgroup analyses of the moderating effects of neuroticism
Given the significant moderation effects of neuroticism, we further explored their context-dependency by conducting stratified analyses for effects that were significant in the prior models (Supplementary Table 1). These results suggest differential moderating patterns across caregiving contexts. Specifically, the amplifying effect of neuroticism on the association between global CDR and ZBI was more pronounced in female (B [SE]=14.202 [4.554], p=0.002) and non-spouse caregivers (B [SE]=10.571 [3.987], p=0.009).
Exploratory analyses of the moderating effects of personality domains other than neuroticism
We conducted exploratory analyses for the remaining personality domains (Supplementary Table 2). Agreeableness significantly moderated the association of global CDR with NPI-D (B [SE]=-3.437 [1.728], p=0.048). Specifically, the effect of global CDR on NPI-D was weaker among caregivers with higher agreeableness (mean +1SD; B [SE]=3.946 [1.055], p=0.002) compared to those with lower agreeableness (mean -1SD; B [SE]=7.273 [1.145], p<0.001). No significant moderating effects were observed for the other personality domains.
DISCUSSION
The present study showed that neuroticism significantly amplified the relationship between cognitive impairment severity (global CDR) and both caregiving burden (ZBI) and caregiver distress (NPI-D). Specifically, as cognitive impairment severity worsened, caregivers with higher levels of neuroticism experienced a steeper increase in their burden and distress compared to those with lower levels of neuroticism. These findings suggest that the caregiving impact of care-recipients’ cognitive decline is not determined solely by carerecipient clinical features, but is substantially conditioned by caregivers’ personality traits, particularly neuroticism, which shapes how clinical stressors are appraised and translated into burden and distress.
The findings for overall caregiving burden revealed that neuroticism significantly moderated caregivers’ responses to increasing cognitive impairment severity. Neuroticism demonstrated a significant amplifying effect, a finding that aligns with its established links to maladaptive emotional tendencies and stress-coping styles [30-32]. Compared with caregiving for other medical conditions, caregiving for individuals with cognitive impairment involves a higher likelihood of unpredictable and challenging situations [33], and these demands tend to intensify as global CDR increases. Under such circumstances, caregivers with high levels of neuroticism tend to appraise these unpredictable caregiving stressors as more threatening and may employ less effective coping strategies [32].
A similar pattern of moderation was observed for NPI-D. Consistent with the pattern observed for overall caregiving burden, neuroticism also significantly amplified the caregiver distress in response to increasing cognitive impairment severity.
Importantly, our analyses revealed a differential pattern whereby neuroticism significantly moderated the associations of global CDR with both ZBI and NPI-D, whereas no such moderation was observed when neuropsychiatric symptoms were indexed by NPI. This pattern suggests that neuroticism’s moderating influence is more pronounced when caregiving stressors reflect broader and cumulative functional decline, as captured by global CDR. Given the profound impact of behavioral disturbances on caregiver stress [15], it is highly plausible that neuropsychiatric symptoms represent a clinical stressor powerful enough to overwhelm individual differences in caregiver personality traits. Methodologically, the structural shared variance between the concurrently measured NPI and NPI-D, as well as the wider variance of NPI compared to the global CDR, may have limited the statistical power to detect significant interaction effects. Collectively, the inclusion of both measures provided a more nuanced understanding of how personality traits differentially shape caregiving experiences across multiple dimensions of cognitive impairment [27,34].
The amplified effects observed in subgroups such as women, non-spouses, and cohabiting caregivers are likely attributable to differences in stress exposure, social role expectations, and access to psychological and social resources. For instance, female caregivers and those with greater daily contact with the care-recipient often face higher emotional and practical caregiving demands, which may render their adaptive or maladaptive personality traits more impactful in shaping their responses to cognitive impairment-related stressors. This is consistent with the stress process model, which posits that caregiver characteristics and contextual factors interact to influence the effects of caregiving stress [35]. Previous literature supports these subgroup differences, linking them to gender-role socialization and the intensity of caregiving involvement [36-40].
Beyond neuroticism, exploratory analyses revealed meaningful moderating patterns for other personality domains. Agreeableness showed a significant buffering effect on NPI-D. Caregivers high in agreeableness, characterized by empathy and effective emotional regulation, may better interpret neuropsychiatric symptoms as disease-driven rather than personally directed, reducing emotional strain [30,41]. Although openness did not reach statistical significance, its trend toward a protective role aligns with prior research linking openness to cognitive flexibility, which may help caregivers adapt to unpredictable caregiving situations [42]. Overall, these findings suggest that while neuroticism amplifies burden and distress, traits such as agreeableness and openness may foster resilience through more adaptive emotional and cognitive coping processes [30,41,42]. Further studies with larger sample sizes are needed to clarify and validate the moderating roles of other personality domains.
These results strengthen the case for developing personality-informed support for caregivers of individuals with cognitive impairment. Such a tailored approach could aim to mitigate the risks associated with certain traits, like neuroticism. Importantly, the present study systematically examined both care-recipient clinical factors and caregiver personality characteristics. In a relatively large caregiver sample, we further captured multiple dimensions of caregiving-related outcomes, including overall caregiving burden and symptom-specific caregiver distress. Integrating these psychological insights into clinical practice may guide healthcare professionals in developing more nuanced and effective strategies to promote adaptive coping among caregivers.
Despite the strengths of our study, several limitations should be acknowledged. First, the cross-sectional design precludes causal inferences. Although personality traits are relatively stable in adulthood [43], longitudinal studies are needed to clarify potential bidirectional relationships between caregiving burden and personality expressions over time. Second, this study did not assess potentially modifiable mediators, such as emotion regulation strategies or perceived social support. Incorporating these factors in future research would provide a more comprehensive understanding of how personal resources buffer the impact of personality on burden [42]. Finally, the scoring of the clinician-rated CDR relies partly on caregiver reports, introducing a potential for respondent bias, such as inflated severity reporting by caregivers who are high in neuroticism. While specialists mitigate this confounding effect by synthesizing comprehensive clinical evaluations, the possibility that subjective caregiver perceptions influence the assessment cannot be entirely ruled out.
In addition to these limitations, future research should examine whether psychological interventions are differentially effective depending on caregiver personality profiles. Although general psychosocial interventions are known to reduce caregiver burden [44], investigating whether personality-informed or tailored approaches yield superior outcomes would provide important clinical implications. Such work could strengthen the rationale for integrating psychological profiling into routine caregiver support.
Supplementary Materials
The Supplement is available with this article at https://doi.org/10.30773/pi.2026.0014.
Exploratory subgroup analysis: moderating effect of personality traits on the association between global CDR and caregiving-related outcomes
Moderating effects of caregiver personality traits other than neuroticism on caregiving-related outcomes
Notes
Availability of Data and Material
The data that support the findings of this study are not publicly available due to ethical and privacy restrictions but are available from the corresponding author upon reasonable request.
Conflicts of Interest
The authors have no potential conflicts of interest to disclose.
Author Contributions
Conceptualization: So Yeon Jeon. Data curation: So Yeon Jeon. Formal analysis: Bo Lim Lee, ChanGyung Kim, Jihye Heo, Dongheon Lee. Funding acquisition: So Yeon Jeon. Investigation: Bo Lim Lee, ChanGyung Kim, So Yeon Jeon. Methodology: Dongheon Lee, So Yeon Jeon. Supervision: So Yeon Jeon. Visualization: Jihye Heo. Writing—original draft: Bo Lim Lee, ChanGyung Kim. Writing—review & editing: all authors.
Funding Statement
This work was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Education (grant number RS-2023-00210380), and by the Korean Ministry of Science and ICT (grant number RS-2026-25476550).
Acknowledgments
The authors thank all the participants and their families for their participation in this study. We would also like to thank the research assistants and study staff.