Currently, over 5.8 million Americans live with Alzheimer disease and related dementias. Over 11 million Americans serve as unpaid caregivers for persons living with dementia (PLWDs; Alzheimer's Association, 2021). Persons living with dementia and their caregivers are sociodemographically and situationally diverse as are the causes and stages of the illnesses with which the PLWD present in primary care settings. Dementia management is complex, including due to PLWDs' declining ability to care for self, which interferes with timely management of comorbidities (Moga et al., 2017). Likewise, dementia is associated with functional impairment (McKhann et al., 2011), neuropsychiatric symptoms (Phan et al., 2019), comorbidities (Poblador-Plou et al., 2014), and institutionalization (Afram et al., 2014). Despite the emergence of several innovative dementia care programs in the United States in the past decade (Clevenger et al., 2018; LaMantia et al., 2015), mainstream primary care settings remain the principal sites for PLWDs' health care (Callahan et al., 2009; Samus et al., 2018).

Quality dementia care is difficult to find in mainstream health care. Challenges include limited time for office visits and clinicians' insufficient knowledge about dementia and resources for PLWD (e.g., social work). Persons living with dementia, with their deteriorating ability to communicate, require more time during the office visits (Hinton et al., 2007). Furthermore, behavioral and psychological symptoms of dementia (e.g., agitation, aggression) are the key reason why PLWD are a "poor fit" for mainstream outpatient settings (Hinton et al., 2007, p. 1488). Health care services for PLWD are fragmented and characterized by inadequate communication (Martin et al., 2020). This results in PLWDs' unmet needs (Miranda-Castillo et al., 2010), poor outcomes, and increasing health care costs (Samus et al., 2014). With no uniform dementia care guidelines, finding a setting with knowledgeable clinicians and an appropriate setup for dementia care is "a matter of luck" (Borson & Chodosh, 2014, p. 396).

In contrast, health care provided by innovative, person-centered, multidisciplinary dementia care teams is recommended by the Alzheimer's Association (Fazio et al., 2018). In these multidisciplinary teams, each staff member must be knowledgeable about dementia and understand the roles of other staff to coordinate care, ensuring that PLWDs' concerns are addressed by staff with appropriate expertise. Furthermore, optimal dementia care presumes continuity of care, with designated staff responsible for PLWD during transitions between care settings and relaying clinical information to staff at the receiving end of the transition (Fazio et al., 2018). In recognition that it is impossible to supply sufficient dementia care specialists for an increasing population of PLWD, experts recommend adopting key principles of dementia-capable health care in existing primary care settings. These principles presume multidisciplinary collaboration. Specifically, more engaged, deliberate inclusion of the nursing and social work staff who are already employed in primary care settings is advised. These staff, in collaboration with primary care providers, may coordinate care, address caregiver burden, and manage PLWDs' neuropsychiatric symptoms. The ultimate goals of transforming health care into dementia-capable care are to improve PLWDs' health outcomes and prevent emergencies (Borson & Chodosh, 2014).

One example of a dementia-capable health care program is the Integrated Memory Care Clinic (IMCC, hereafter, Clinic)-an innovative multidisciplinary dementia care program-the setting in which the current study was conducted. A cohort of newly enrolled PLWD and their caregivers at the Clinic was examined. The purpose of this study was to identify characteristics associated with: (1) PLWDs' neuropsychiatric symptoms and quality of life; and (2) distress from neuropsychiatric symptoms and well-being among their caregivers. By identifying these characteristics, we aimed to provide primary care clinicians with information about the heterogeneity they are likely to encounter as new PLWD and their caregivers enter their practices. With scarce time and resources in busy primary care settings, it may be useful for clinicians to have insights about the characteristics associated with PLWDs' neuropsychiatric symptoms and caregivers' distress from these symptoms and diminished well-being.

Methods

Design

We report data from a cross-sectional single-group examination of PLWD and caregivers collected during new patient enrollment at the Clinic, a specialty care setting for PLWD (Cellar & Clevenger, 2018; Clevenger et al., 2018). This examination is one in the series of studies of PLWDs' and caregivers' outcomes (Clevenger et al., 2018; Kovaleva et al., 2020) and caregivers' experience (Kovaleva et al., 2021) in this innovative dementia care program. The study was approved by the Emory University Institutional Review Board.

Setting

The Clinic is an accredited patient-centered medical home founded and led by advanced practice registered nurses (APRNs). Advanced practice registered nurses, supported by a geriatrician, a registered nurse, a social worker, and office staff, deliver dementia care and primary care concurrently throughout the PLWDs' lifetime or until their institutionalization. The Clinic prioritizes optimization of quality of life and well-being of PLWDs and caregivers. Persons living with dementia, to the degree they are able, and their caregivers partner with APRNs in clinical decision-making. Symptoms are managed aggressively within a palliative care framework. Discussions on goals of care and advance care planning are prioritized. Persons living with dementia and caregivers come to the Clinic through self-referral or physician referral and arrive with all types and stages of dementia. Caregivers are usually an integral part of the Clinic visits, supplementing PLWDs' health history and reporting on PLWDs' function and neuropsychiatric symptoms. Detailed design principles of this Clinic are described elsewhere (Cellar & Clevenger, 2018; Clevenger et al., 2018).

Sample

Caregivers were recruited using flyers at the IMCC and by Clinic staff introducing the study to potentially eligible participants. Unpaid English-speaking caregivers, at least 18 years old, caring for community-dwelling PLWD, who enrolled in the Clinic no earlier than 90 days before data collection were eligible to participate. Power analysis was conducted in G*Power (Faul et al., 2007), yielding a sample of 45 caregivers based on 80% power and 5% significance level. The final target sample of 54 accounted for 20% attrition. Of the 189 caregivers invited, 49 caregivers verbally consented and completed the baseline assessment (Kovaleva et al., 2020).

Data collection

Data for all instruments other than the Neuropsychiatric Inventory-Questionnaire (NPI-Q; Kaufer et al., 2000) were collected from caregivers via telephone (as opposed to via in-person visits) to save participants' time. Caregivers received a $15 gift card for completing the questionnaires. Clinic staff administered the NPI-Q to caregivers on enrollment and semiannually: these data were retrieved from electronic health records (EHR). All data were collected from caregivers, not PLWD. Caregivers reported on their own psychological well-being and health status and evaluated their PLWDs' quality of life and frequency and severity of neuropsychiatric symptoms. Additionally, caregivers evaluated the distress they experienced from their PLWDs' neuropsychiatric symptoms.

Measures and research instruments

Independent variables

Independent variables included the following: age, gender, race, education (less than college or at least a college degree), employment (unemployed or employed), relationship to PLWD (spouse or adult child), caregiving duration, receiving help with caregiving (yes or no), PLWD's age, and years since dementia onset. Dementia type and PLWDs' comorbidities were retrieved from the EHR. The Charlson Comorbidity Index was calculated (Charlson et al., 1994).

Dependent variables

Variables related to neuropsychiatric symptoms and PLWDs' and caregivers' well-being were ascertained using multiple measures. Caregivers reported on their well-being: perceived stress (Perceived Stress Scale; Cohen et al., 1983), depressive symptoms (Center for Epidemiologic Studies-Depression; Andresen et al., 1994; Radloff, 1977), caregiver burden (Zarit Burden Interview; Bedard et al., 2001; Zarit et al., 1980), anxiety (Patient-Reported Outcomes Measurement Information System-Anxiety; PROMIS, n.d.), health status (Short Form Health Survey; Ware & Sherbourne, 1992), and distress from PLWDs' neuropsychiatric symptoms (Neuropsychiatric Inventory-Questionnaire; Kaufer et al., 2000). Additionally, caregivers reported on PLWDs' quality of life (Quality of Life-Alzheimer's Disease; Logsdon et al., 2002) and severity of neuropsychiatric symptoms (Neuropsychiatric Inventory-Questionnaire; Kaufer et al., 2000). The Healthy Aging Brain Care-Monitor was used to assess the frequency of PLWDs' cognitive, behavioral and mood, and functional symptoms, and caregivers' stress (Monahan et al., 2012). In-depth description of the instruments with scoring and psychometric properties is provided elsewhere (Kovaleva et al., 2020).

Data analysis

Analyses were run in IBM SPSS v. 26 (IBM, 2018). The sample was characterized using descriptive univariate analyses. Pearson correlations were used to assess the strength of the associations of the PLWD and caregiver characteristics with neuropsychiatric symptoms and caregivers' well-being. Several continuous variables were transformed to meet normality assumptions of Pearson correlation. Several categories of the nominal variables were combined to create dichotomous variables due to small cell sizes. These included race (Black or White), dementia type (Alzheimer or non-Alzheimer), caregiver education (less than college or college and beyond), and relationship configuration (spouse or adult child). Statistically significant correlations (p < .05) were converted to Cohen d statistics for ease of interpretation.

Results

Sample

Forty-nine caregivers completed the questionnaires (35 women, mean age 63.8 years, 26 spousal caregivers, 14 Black). Of these 49 caregivers, 25 had postgraduate/professional education and 13 had college degrees. Thirty-nine caregivers were employed outside the home. Persons living with dementia were, on average, 77.2 years old. Thirty-two PLWD had Alzheimer disease with a median time since dementia onset of 3 years. Other dementia types included mixed, vascular, frontotemporal, dementia with Lewy bodies, and other dementias. Persons living with dementia had a median of five chronic conditions besides dementia (range 1-7) and a median Charlson Comorbidity Index of 5 (range 3-10). The most frequent comorbidities were hypertension (n = 27), musculoskeletal conditions (n = 24), and hyperlipidemia (n = 23). Forty-three caregivers lived with their PLWD. In eight of those cases, caregivers also lived with others in addition to the PLWD. Forty caregivers received assistance with their caregiving, 16 of the 40 paid for the assistance. Common sources of unpaid help were caregivers' daughters (n = 14) and sisters (n = 10). Nineteen caregivers were responsible for caring for others in addition to caring for PLWD (e.g., spouse, children).

To characterize the sample in greater detail, Pearson correlation and Chi-square tests were run between all baseline sociodemographic and clinical variables (Table 1). Statistically significant correlations are described below.

Table 1. Correlation matrix between sociodemographic and clinical variables at baseline (

Caregiver age

On average, male caregivers (n = 14, mean age 72 years, SD 10.3) were older than female caregivers (n = 35, mean age 61 years, SD 10.7, p = .01, Cohen d 1). Employed caregivers were younger (n = 19, mean 58 years, SD 11.9) than unemployed caregivers (n = 30, mean 67.5 years, SD 10.1, p = .004, Cohen's d 0.88). Spousal caregivers (n = 26, mean 72.6 years, SD 6.6) were older than adult children caregivers (n = 23, mean 54 years, SD 7.6, p < .001, Cohen's d 2.67). Caregivers who were responsible for at least one other person in addition to the PLWD were younger (n = 19, mean 59 years, SD 8.6) than caregivers who did not have such responsibility (n = 30, mean 67 years, SD 12.5, p = .019, Cohen's d 0.72).

Caregiver gender, relationship configuration, and employment

Male caregivers were more likely to care for a spouse (n = 11) than for a parent (n = 3) ([chi]²(1) = 5.12, p = .024, Cohen's d 0.68). Spousal caregivers cared for PLWD who had a lower Charlson Comorbidity Index (n = 26, mean 4.1, SD 0.07) compared with PLWD whose caregivers were adult children (n = 23, mean 5.4, SD 0.16, p = .005, Cohen's d 0.87). Spousal caregivers were more likely to be unemployed (n = 20) compared with adult children caregivers (n = 10) ([chi]²(1) = 5.75, p = .016, Cohen's d 0.73).

Caregiver race

Because data were collected from caregivers, only caregiver race was ascertained. To the authors' best knowledge, however, all dyads were racially concordant. Black caregivers (n = 14, mean 59.1 years, SD 8) were younger than White caregivers (n = 32, mean 66.7 years, SD 12, p = .040, Cohen's d 0.64). Black caregivers were also more likely to care for parents (n = 11) than for spouses (n = 3) ([chi]²(1) = 8.79, p = .003, Cohen's d 0.97). Persons living with dementias whose caregivers were Black had higher Charlson Comorbidity Indices (mean 5.6, SD 0.18) than PLWD whose caregivers were White (mean 4.3, SD 0.1, p = .017, Cohen's d 0.75). Analogously, PLWD whose caregivers were Black had more comorbidities (mean 7.1, SD 0.88) compared with PLWD whose caregivers were White (mean 4.3, SD 0.69, p = .037, Cohen's d 0.65).

Persons living with dementias' age, Charlson comorbidity index, and time since dementia onset

Spousal caregivers cared for PLWD who were younger (n = 26, mean 73.7 years, SD 6.8) than PLWD whose caregivers were children (n = 23, mean 81.1 years, SD 9, p = .002, Cohen's d 0.96). As PLWDs' age increased, their Charlson Comorbidity Index also increased (r = 0.8, p < .001, Cohen's d 2.65), as did their number of comorbidities (r = 0.47, p = .001, Cohen's d 1.07). Higher Charlson Comorbidity Index was associated with PLWDs' higher number of comorbidities (r = 0.57, p < .001, Cohen's d 1.39). Longer time since dementia onset was associated with increased time caregivers had been caring for their PLWD (r = 0.72, p < .001, Cohen's d 2.08).

Results for caregivers

Several PLWD-related variables were strongly associated with caregivers' diminished well-being (Supplementary Table 2, Supplemental Digital Content 2, http://links.lww.com/JAANP/A142). The strongest effect was observed for the time since dementia onset: the longer the PLWD had dementia, the greater were caregivers' self-reported role limitations due to caregivers' physical health and emotional problems (Cohen's d 0.82 and 0.75). Additionally, non-Alzheimer dementia was associated with higher caregiver burden (Cohen's d 0.69). Finally, higher Charlson Comorbidity Index scores were linked to higher stress among caregivers (Cohen's d 0.63).

Several caregiver characteristics were linked to caregivers' higher distress from their PLWDs' neuropsychiatric symptoms. The strongest effects were observed for caregiver's race: Black caregivers reported higher total distress from PLWDs' neuropsychiatric symptoms (Cohen's d 1.39) and, more specifically, from PLWDs' delusions, disinhibition, and agitation (Cohen's d 1.64, 1.07, and 1.03). Strong effects were also observed for caregiver's age: younger caregivers reported higher distress from their PLWDs' sleep disturbance, apathy, appetite and eating disorders, and higher total distress from PLWDs' all symptoms (Cohen's d 1.5, 1.22, 1.01, and 1.16). Additionally, strong effects were observed for the caregiver-PLWD relationship: adult children caregivers reported higher distress from their PLWDs' sleep disturbance and higher total distress from PLWDs' symptoms (Cohen's d 1.42 and 1.05). Finally, employed caregivers had higher distress from their PLWDs' depression, apathy, appetite and eating disorders, and total distress from all neuropsychiatric symptoms (Cohen's d 0.84, 0.79, 0.68, and 0.76).

Results for PLWD as reported by their caregivers

Several PLWD-related variables were strongly associated with higher symptom burden and lower quality of life for PLWD as reported by their caregivers (Supplementary Table 3, Supplemental Digital Content 3, http://links.lww.com/JAANP/A143). Non-Alzheimer dementia and longer time since dementia onset were associated with PLWDs' more severe sleep disturbance (Cohen's d 0.7 and 0.65). Longer time since dementia onset was also associated with PLWDs' lower quality of life (Cohen's d 0.67). Finally, more comorbidities and a higher Charlson Comorbidity Index were associated with more neuropsychiatric/dementia symptoms (Cohen's d 0.65 and 0.71).

Several caregiver-related variables were associated with PLWDs' higher frequency and severity of neuropsychiatric symptoms as perceived by caregivers. Black caregivers reported higher total symptom severity and higher severity of delusions, agitation, and disinhibition for their PLWD (Cohen's d 1.27, 1.63, 1.17, and 1.1). Additionally, Black caregivers reported that their PLWD had more neuropsychiatric symptoms (Cohen's d 1.34). Regardless of race, female caregivers reported higher severity of their PLWDs' sleep disturbance and appetite and eating disorders compared with male caregivers (Cohen's d 0.97). Younger caregivers reported higher severity of their PLWDs' sleep disturbance and higher total symptom severity (Cohen's d 1.25 and 1.08). Adult children caregivers reported higher severity of their PLWDs' sleep disturbance compared with those caring for their spouses (Cohen's d 1.04). Finally, employed caregivers reported higher severity of their PLWDs' apathy and depression and higher total symptom severity (Cohen's d 0.72, 0.7, and 0.66).

Discussion

This study is one in a series of investigations of the Clinic: its design, implementation, PLWDs' and caregivers' outcomes, and caregivers' experience (Cellar & Clevenger, 2018; Clevenger et al., 2018; Kovaleva et al., 2020, 2021; Kovaleva et al., 2020, 2021). Future explorations of PLWDs' and caregivers' outcomes in this care program are planned and will add to the body of literature on the contributions and results of the innovative multidisciplinary dementia care programs (Boustani et al., 2011; LaMantia et al., 2015; Noel et al., 2017; Reuben et al., 2013).

The findings from this study illustrate the considerable differences among a group of 49 dyads as they entered a combined dementia and primary care practice as new clients. It highlights a cluster of key elements that characterize the heterogeneity of the group. Consistent with expectations, characteristics, including non-Alzheimer dementia, longer time since dementia onset, more comorbidities, higher Charlson Comorbidity Index, and caregivers' female gender showed associations with PLWDs' higher symptom burden and lower quality of life. These characteristics also showed associations with caregivers' higher distress from their PLWD's neuropsychiatric symptoms and diminished health status and well-being. Less expected were higher symptom burden and lower quality of life among PLWD whose caregivers were Black, younger, employed, and/or adult children. These results identify characteristics of PLWD-caregiver dyads that may be easily overlooked but may represent red flags for distress. Ignoring clients who seem to be managing well (e.g., because of their younger age or cultural family values) could perpetuate distress for PLWD and caregivers alike. Of practical importance, the following key characteristics were compiled into a checklist for clinicians to assess PLWD and caregivers and tailor their care plans (Supplementary Figure, Supplemental Digital Content 1, http://links.lww.com/JAANP/A141).

Race

Black caregivers rated their PLWDs' neuropsychiatric symptoms as more severe and their distress from these symptoms higher compared with White caregivers. Race also most frequently accounted for the variables associated with PLWDs' higher symptom severity. Our findings accord with previous research documenting the prevalence of neuropsychiatric symptoms among Black PLWD (Schoos et al., 2004). It is important that, however, this result contradicts a well-established pattern in the literature asserting better coping and psychological well-being among Black caregivers compared with Whites (Skarupski et al., 2009). This pattern may be based on an incorrect stereotype that Black family values both underscore caregiving responsibility and buffer caregiving stress (Pharr et al., 2014). Particularly with a long history of disparities in care (Alzheimer's Association, 2021; Chin et al., 2011; Dilworth-Anderson et al., 2012; Zuckerman et al., 2008), clinicians should bypass this stereotype and ask Black caregivers specifically about PLWDs' neuropsychiatric symptoms and their own associated distress. More importantly, clinicians should offer concrete and feasible action items for managing Black PLWDs' neuropsychiatric symptoms and associated caregivers' distress and regularly follow-up on the implementation of this care plan.

It is important that, PLWD whose caregivers were Black had statistically significantly more comorbidities and higher Charlson Comorbidity Index. This may indicate that Black PLWD entered the Clinic as more morbid compared with White PLWD. This finding accords with the well-documented evidence that older Black adults' health is worse compared with that among, for example, Whites and Hispanics. Specific examples of disparities include worse cardiometabolic health (Mitchell et al., 2019) and more prevalent underinsured status (Sohn, 2017). Moreover, between 1999 and 2018, Blacks experienced increases in the prevalence of hypertension, diabetes, stroke, and asthma (Odlum et al., 2020). Clinicians, therefore, ought to manage dementia in the context of overall high multimorbidity among Black PLWD.

Caregiver's gender, age, relationship configuration, and employment

Female, younger, adult children of PLWD and employed caregivers reported higher severity of their PLWDs' neuropsychiatric symptoms and higher distress from these symptoms. Likewise, we found that those caregivers who cared for at least one other person in addition to PLWD were statistically significantly younger than those who cared only for PLWD-likely indicating younger adult children caregivers. Furthermore, we found that adult children caregivers were statistically significantly more likely to care for PLWD who are older and who have higher Charlson Comorbidity Index.

The finding of younger caregivers having higher distress differs from reports that spouses had more severe depressive symptoms than adult children (Pinquart & Sorensen, 2011). The reason behind this discrepancy is not known. Our results, however, coincide with more recent findings of female caregivers' reporting higher burden and worse mood (Ma et al., 2018; Xiong et al., 2020). Also, our findings parallel those in a systematic review that female and adult children caregivers are more likely to have depressive symptoms compared with spousal caregivers (Watson et al., 2019). Adult children cared for PLWD (their parents) whose Charlson Comorbidity Index was higher compared with that of PLWD whose caregivers were spouses. Caring for PLWD with a higher Charlson Comorbidity Index indicates that other illnesses may contribute to PLWDs' symptoms (e.g., agitation caused by pain due to musculoskeletal conditions). Thus, it is important to holistically evaluate PLWDs' health for other comorbidities adding to symptom severity because some comorbidities may be controlled pharmacologically or nonpharmacologically (e.g., pain control). Attention to these caregivers is crucial because younger adult children caregivers may be erroneously presumed to be doing well due to their younger age, not necessarily co-residing with PLWD, and having own family.

Weaker effect sizes were observed for the relationship between caregivers' employment, their PLWDs' higher symptom severity, and caregivers' higher distress from neuropsychiatric symptoms. Additionally, we found that adult children caregivers were statistically significantly more likely to be employed, highlighting their other responsibilities, beyond caregiving. Clinicians might proactively ask about caregiver stress in caregivers who are juggling both caregiving and employment. The damaging effect of caregiving on caregivers' employment is well-documented (Alzheimer's Association, 2021). Caregivers' distress may be compounded by the uncertainty about how to provide care to their PLWD while anticipating threats to their employment, including financial stability and nonmonetary benefits of employment (e.g., prestige, social status, socialization outside of home, sense of belonging to a professional group). Support for these caregivers may range from validation of their feelings and experiences to connecting them to social workers who may advise them regarding financial and respite assistance. Caregivers may benefit from reappraising their financial situation to allow for supplementary paid help. Clinicians may recommend caregivers take a more assertive stance with other family members who may not provide hands-on or financial assistance about financial and human resources that caregiving requires.

Time since dementia onset

Not surprisingly, the length of time since dementia onset seems to have an important bearing on PLWDs' and caregivers' well-being. Longer time since dementia onset was associated with PLWDs' lower quality of life, a self-evident finding given the progressive nature of dementia (Alzheimer's Association, 2021). More time since dementia onset was also associated with more severe sleep disturbances in PLWD. This result is expected because sleep disturbance is prevalent among PLWD and effective treatments are lacking (Kinnunen et al., 2017). These results are also linked to wear-and-tear effects on caregivers' physical and emotional health and overall stress.

Comorbidities

Our results indicate that having more comorbidities and a higher Charlson Comorbidity Index is associated with a higher frequency of neuropsychiatric symptoms. Deteriorating ability of PLWD to communicate may lead to an inability to express their needs, triggering PLWDs' frustration and agitation and complicating management of comorbidities. A clinical implication is to create a comprehensive care plan where no chronic conditions or symptoms are overlooked and the effect of dementia on all these conditions and symptoms is evaluated.

Dementia type

In our study, PLWD with non-Alzheimer dementia had higher symptom severity compared with those with Alzheimer disease. This finding fits with previous research: persons with rarer dementia types frequently have higher symptom severity (Russ et al., 2015). Non-Alzheimer dementias are less prevalent (Alzheimer's Association, 2021), so clinicians may be less familiar with managing them. For example, caregivers of persons living with frontotemporal dementia report difficulties finding health care professionals knowledgeable about this condition (Rasmussen et al., 2019). A clinical implication is to provide caregivers with resources specific to their PLWD's rarer dementia type (e.g., caregiver education dedicated to a particular dementia type, referral to a specialist in the respective area).

Our findings illustrate the variety of patterns of higher PLWD symptom burden and diminished caregiver well-being and offer areas of assessment and treatment. The principal implication of these findings is that a more nuanced understanding of PLWDs' and caregivers' characteristics is necessary. More precise understanding of the interplay among PLWDs' neuropsychiatric symptoms, their caregivers' well-being, and PLWDs' and caregivers' sociodemographic and clinical characteristics may help enrich intake assessment processes. More individualized care plans may help attain patient-centered care and, most importantly, allow clinicians to intervene when clients are in distress, even if this distress is concealed by the individuals themselves or assumptions about them. Managing PLWDs' neuropsychiatric symptoms is essential because neuropsychiatric symptoms are extremely challenging for PLWD and their caregivers, yet they are often undertreated (Kales et al., 2014).

Such assessments might also prompt providing caregiver education about common neuropsychiatric symptoms (e.g., apathy, depression) and rarer ones (e.g., elation/euphoria). Clinicians may educate caregivers about the difference between irreversible cognitive changes in dementia vs. modifiable neuropsychiatric symptoms. Caregivers will benefit from knowledge about the potential causes, moderators, and pharmacologic and nonpharmacologic management of neuropsychiatric symptoms, and about the inevitable progression of dementia despite caregivers' best efforts (Desai et al., 2012).

This study has several important limitations. Retrieving data for the Neuropsychiatric Inventory from the EHR was necessary to decrease participant burden but it resulted in missing data. Only associative relationships can be traced in this cross-sectional study. Results from this small convenience sample drawn from a single dementia care program in an academic health center must be interpreted with caution and may not be generalizable to mainstream health care settings where most PLWD receive health care. Similarly, the results may not be applicable to care settings that differ starkly from an academic health center (e.g., dyads living in rural areas, those receiving health care in safety-net health care systems, Veterans, ethnic minorities, and those who do not speak English). Moreover, over 75% of the caregivers had college or postgraduate degrees and over 80% received paid or unpaid in-home assistance. Thus, this sample differs from caregivers who have fewer resources.

The strengths of the current study, however, are in the detailed examination of the sociodemographic and clinical factors associated with PLWDs' neuropsychiatric symptoms and their caregivers' distress. Despite the apparent homogeneity of the sample, the finding of heterogeneity of PLWDs' symptom burden and PLWDs' and caregivers' well-being highlights the importance of providing timely and precise education and support to this population. Exploring the associations between these variables highlights relationships that may not be apparent to clinicians intuitively. This analysis underscores the importance of attention to caregivers who may not seem overly distressed. Likewise, the results of this study stress the importance of assessment of PLWDs' and caregivers' outcomes using established instruments. A similar practice may be adapted in health care settings where PLWD and their caregivers are served, as opposed to solely asking open-ended questions about caregivers' work or erroneously assuming no issues exist for these caregivers.

Conclusion

In this study, we identified key characteristics associated with more severe and frequent neuropsychiatric symptoms among PLWD and higher distress from these symptoms among their caregivers. These results challenge stereotypes and assumptions about caregivers. This diagnostic cross-sectional view of PLWDs' and caregivers' most problematic areas may be useful to clinicians who work with families affected by dementia in mainstream health care settings. The findings may be used as a checklist by clinicians in busy clinical environments to better address caregivers' needs in a precise, personalized manner that can make a significant difference for caregivers who too often mask their own distress.

Acknowledgements:The authors are grateful to all caregivers who participated in this study despite their numerous responsibilities. The authors would like to thank Tyler Finkle and Joshua Schwartz who helped with caregivers' recruitment and Jean Harrell who assisted with the study management. The authors are thankful to the staff of the Integrated Memory Care Clinic: Carolyn Clevenger, Janet Cellar, Debbie Gunter, Alison Schlenger, Laura Medders, and Heidi Wells. The principal investigator is indebted to Janet Cellar, Elizabeth Corwin, Rebecca Gary, Melinda Higgins, Bonnie Jennings, Ann Rogers, Kathryn Wood, and Kate Yeager for the letters of recommendation for grant applications that funded this study. The authors would like to acknowledge the authors of the Healthy Aging Brain Care Monitor for the permission to use this instrument in this study: Patrick Monahan, Catherine A. Alder, Babar A. Khan, Timothy Stump, and Malaz A. Boustani.

References