Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 May 1.
Published in final edited form as: Int Psychogeriatr. 2023 May 29;35(11):653–663. doi: 10.1017/S1041610223000467

Malnutrition and Neuropsychiatric Symptoms in Dementia: The Cache County Dementia Progression Study

Kaitlyn Kauzor 1, Mikaela Drewel 1, Hector Gonzalez 1, Gail B Rattinger 2, Alexandra G Hammond 1, Heidi Wengreen 3, Constantine G Lyketsos 4, JoAnn T Tschanz 1,5
PMCID: PMC10592578  NIHMSID: NIHMS1898613  PMID: 37246509

Abstract

Objectives:

Among people with dementia, poor nutritional status has been associated with worse cognitive and functional decline, but few studies have examined its association with neuropsychiatric symptoms (NPS). We examined this topic in a population-based sample of persons with dementia.

Design:

Longitudinal, observational cohort study.

Setting:

Community

Participants:

Two hundred ninety-two persons with dementia (71.9% Alzheimer’s Disease, 56.2% women) were followed up to six years.

Measurements:

We used a modified Mini-Nutritional Assessment (mMNA) and the Neuropsychiatric Inventory (NPI) to evaluate nutritional status and NPS, respectively. Individual linear mixed effects models examined the associations between time-varying mMNA total score or clinical categories (malnourishment, risk for malnourishment, or well-nourished) and NPI total score (excluding appetite domain) or NPI individual domain or cluster (e.g., psychosis) scores. Covariates tested were dementia onset age, type, and duration, medical comorbidities, sex, APOE genotype, and education.

Results:

Compared to the well-nourished, those at risk for malnourishment and those malnourished had higher total NPI scores [b (95% CI) = 1.76 (0.04, 3.48) or 3.20 (0.62, 5.78), respectively], controlling for significant covariates. Higher mMNA total score (better nutritional status) was associated with lower total NPI [b (95% CI) = −0.58 (−0.86, −0.29)] and lower domain scores for psychosis [b (95% CI) = −0.08 (−.16, −.004)], depression [b (95% CI = −0.11 (−0.16, −0.05], and apathy [b (95% CI = −0.19 (−0.28, −0.11)].

Conclusions:

Worse nutritional status is associated with more severe NPS. Dietary or behavioral interventions to prevent malnutrition may be beneficial for persons with dementia.

Introduction

A decline in nutritional status amongst older adults is common (Dorner, 2010) and may contribute to physical morbidity such as involuntary weight loss, reduced muscle mass/wasting, reduced immune function, slower wound healing, and frailty (Norman et al., 2021). Among persons with dementia, poor nutritional status has been associated with worse cognitive and functional status and mortality (Guerin et al., 2005; Spaccavento et al., 2009; Sanders et al., 2016; Sanders et al., 2018). Recent work also suggests an association with neuropsychiatric symptoms (NPS). For example, malnourished individuals with AD exhibited greater frequency and intensity of NPS over one year in a large, French clinic-based sample (Guerin et al., 2005) and amongst women with mild cognitive impairment or early-stage AD in a Japanese clinic-based sample (Kimura et al., 2019; Kishino et al., 2022). With respect to specific NPS, verbal aggression/emotional disinhibition (Kimura et al., 2019; Kishino et al., 2022), apathy (Spaccavento et al., 2009; Roque et al., 2013; Kimura et al., 2019; Kishino et al., 2022), appetite/feeding behaviors (Roque et al., 2013), hallucinations (Spaccavento et al., 2009, Roque et al., 2013), anxiety (Roque et al., 2013), and sleep (Roque et al., 2013) have been reported in various samples of dementia.

While informative, all the above studies were conducted with individuals recruited from specialty clinics, potentially limiting the generalizability of results to the broader population of persons with dementia who may differ on age, education, health, dementia severity, and other factors [discussed in Tschanz et al., 2011]. Additionally, many of the above cited studies were cross-sectional or with limited duration of follow-up (for example, up to two years). Thus, we examined the association between nutritional status and NPS in a population-based sample of persons with dementia with follow-up duration up to six years. We examined the potential confounding effects of medical/health status at each follow-up in testing our central hypothesis that poorer nutritional status would be associated with greater severity of NPS. This project builds on prior studies reporting the relevance of nutritional status and NPS globally, adding granularity (specific symptoms) and longer follow-up duration in a population cohort, all of which likely enhance the generalizability of results. Furthermore, we controlled for a broad range of medical comorbidities which has been examined in a scant number of papers and with limited scope. We previously published work in this cohort on nutritional status and cognitive and functional decline in dementia progression (Sanders et al., 2016; Sanders et al., 2018), and now extend our analyses to NPS.

Methods

Data Source and Study Sample

The Cache County Dementia Progression Study (DPS) (Tschanz et al., 2011) is a population-based study that examined factors influencing the rate of dementia progression amongst individuals with recent onset dementia identified from the Cache County Study on Memory in Aging (CCSMA). Through the procedures of the CCSMA, 5092 (90%) of the permanent residents of Cache County, Utah underwent up to four triennial waves of dementia screening and ascertainment between 1995 – 2007. Each wave consisted of a multi-staged protocol that commenced with the administration of a brief cognitive screener (Tschanz et al., 2002) and culminated in a clinical assessment (and 18-month follow-up re-assessment), brain neuroimaging, laboratory studies as well as a physician evaluation for those with suspected dementia (Breitner et al., 1999; Miech et al., 2002). Lifetime risk factor interviews inquired about demographic information as well as medical histories in Wave 1 of the CCSMA; updates were obtained in subsequent waves. Genotyping at the Apolipoprotein E (APOE) gene was determined by polymerase chain reaction of DNA from buccal-swab samples (Breitner et al., 1999). Dementia diagnosis was determined by an expert panel of neurologists, neuropsychologists, geriatric psychiatrists, and a cognitive neuroscientist after review of all clinical data. A diagnosis of dementia was assigned based on criteria from the Diagnostic Statistical Manual-III-Revised (DSM-III-R) (American Psychiatric Association, 1987). Assignment of the underlying cause of dementia followed standard research criteria at the time (Breitner et al., 1999). For example, AD diagnosis followed criteria of the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) (McKhann et al., 1984) and vascular dementia (VaD) followed criteria of the National Institute of Neurological Disorders and Stroke and Association Internationale pour la Recherche et l’Enseignement en Neurosciences (NINDS-AIREN) (Román et al., 1993). Age of dementia onset was recorded as the individual’s age when they first met DSM-III-R criteria for dementia.

Surviving individuals identified as incident cases of dementia after the start of Wave 1 were invited to enroll in the DPS (Tschanz et al., 2011). Persons with dementia and their caregivers were visited in their homes approximately every six months where they underwent physical and neurological examinations and neuropsychological assessments, and caregivers provided updated health, nutritional (alternating visits), clinical, functional and NPS information on the participant. All procedures of the CCSMA were approved by the Institutional Review Boards (IRB) at Utah State University, Johns Hopkins University, Duke University, and the University of Washington. The procedures of the DPS were approved by the IRBs of Utah State University and Johns Hopkins University.

Measures

Dependent Variable: Neuropsychiatric Inventory.

The Neuropsychiatric Inventory-12 (NPI-12) (Cummings, 1994) was completed by the participant’s caregiver at baseline and each of the follow-up visits. The NPI-12 assesses NPS across 12 domains which include: delusions, hallucinations, agitation/aggression, depression/dysphoria, anxiety, elation/euphoria, apathy/indifference, disinhibition, irritability/lability, aberrant motor behavior, nighttime behavioral disturbances, and appetite/eating changes (Cummings, 1997). For each symptom endorsed, the caregiver rated both the severity (mild, moderate, marked) and frequency (1 = occasionally, less than once per week; 2 = often, about once per week; 3 = frequently, several times per week but less than every day; 4 = very frequently, once or more per day or continuously) of each symptom. Frequency and severity scores were multiplied to yield an overall severity score (maximum of 12) for each domain, which was summed across all domains to yield a total NPI score (Cummings, 1994). Because the last domain, appetite/eating changes, was associated with the primary independent variable (nutritional status), we excluded this domain in all analyses. Thus, the 11-item total NPI score yielded a maximum of 132 points. To explore specific NPI domains, we considered prior work in the DPS (Vernon et al., 2019) that summed domain scores for co-occurring symptoms [e.g., affective behavior (depression, anxiety and irritability) and psychosis (delusions and hallucinations). We examined NPI symptom intercorrelations and retained those that were moderately correlated (e.g., r = 0.40 or above). Thus, delusions and hallucination severity scores were summed to represent the domain, psychosis (score range 0 – 24 points). Other NPI scores were modeled as individual outcomes (score range 0 – 12) except for elation, which was not examined due to its rare occurrence.

Independent Variable: Modified Mini-Nutritional Assessment.

The Mini-Nutritional Assessment (MNA; Guigoz et al., 1997) is a validated and widely used assessment of nutritional status for older adults adapted for the DPS (Sanders et al., 2016; Sanders et al., 2018). The MNA consists of 18 items categorized into four domains: anthropometric assessment [four items capturing body mass index (BMI), weight loss, and arm and calf circumferences], medical assessment (six items related to mobility, medication use, lifestyle, and psychiatric symptoms), short dietary assessment [eight items about type (e.g., protein, fruits, and vegetables), frequency, and mode of food and fluid intake], and subjective assessment (two items related to self-view of nutritional and health status). With a total of 30 points possible (higher scores representing better nutritional status), the MNA has three cut-off scores for the following clinical categories: malnourishment (less than 17), risk for malnourishment (17 - 23.5), and well-nourished (24 - 30) (Guigoz et el., 1997; Vellas et al., 1999). As previously reported (Sanders et al., 2016), a modified MNA (mMNA) score was determined on alternating visits (i.e., annually) from data gathered from the DPS interviews covering health, medication, nutrition, and adult daily living activities from caregivers and physical examination of participants. MNA items excluded were self-report of nutritional and health status due to questionable reliability and validity from dementia participants. Additionally, skin ulcers and calf-circumference were not assessed. Furthermore, items relating to cognitive status and psychological stress were excluded due to invariance of the former (all participants had dementia) and the outcome (NPS) being modeled with the latter. The adjusted mMNA maximum score was 22 points, with corresponding clinical cut-off scores of malnourishment (less than or equal to 12.5), risk for malnourishment (12.6–17.5), and well-nourished (more than 17.5) (Sanders et al., 2016). The mMNA total score (0 – 22 points) and the clinical categories were entered as time-varying variables in statistical analyses.

Covariates

Variables that were previously found to be associated with nutritional status and NPS were examined as covariates as described below.

Medical Comorbidities

To examine potential confounding from medical comorbidities, we examined overall health rating and number of medical health conditions. A general health rating (GMHR; Lyketsos et al., 1999) was determined by the examining nurse that was based on the physical examination, medication use and the degree to which health conditions were controlled by treatments. A rating of “excellent,” “good,” “fair” or “poor” was assigned. Due to infrequent assignments of “excellent” and “poor,” ratings were combined into “excellent or good” and “fair or poor” health. The GMHR was entered as a time-varying covariate.

Health conditions were queried at each visit from the CCSMA risk factor interviews completed at each wave and subsequent health interviews conducted as part of the DPS visits. A history of the following chronic and acute conditions was queried: thyroid disorders, Parkinson’s disease, seizures, hypertension, high cholesterol, diabetes, arthritis, headache and other painful conditions, cancer, chronic obstructive pulmonary disease, congestive heart failure, head injury, brain surgery, heart attack, stroke, surgeries, and other serious physical illnesses not otherwise queried. We examined the frequencies of the number of physical health-related conditions by visit and created three groups that approximated 20 - 30% of the sample across the visits. Number of health conditions were thus categorized as 0 – 3, 4 – 5, and 6 or more and entered as a time-varying covariate in statistical models.

Other variables

As in prior work with the mMNA (Sanders et al., 2016) and other literature, we tested other covariates including dementia type (Fernández-Martínez et al., 2008; Bjoerke-Bertheussen et al., 2012), age of dementia onset and dementia duration from onset age to baseline assessment (Steinberg et al., 2008), education (Kwak et al., 2020), APOE genotype (Delano-Wood et al., 2008), and sex.

Analyses

Descriptive statistics characterized the sample. Inferential statistics (t-tests for continuous variables and chi-suare tests for categorical variables) were run to examine differences among those included vs. excluded in the sample. Separate linear mixed effects models with random intercepts and slopes were used to examine the association between the primary predictor of nutritional status and trajectories of NPI total score and domain scores (psychosis, agitation/aggression, depression, apathy, anxiety, irritability, aberrant motor behaviors, disinhibition, and nighttime behaviors). Random intercepts accounted for the individual variation in NPI scores at baseline, whereas random slopes accounted for the differences in the rate of change for each participant. In one set of models, we used the continuous, time-varying mMNA total score as the primary predictor and in another set of models, the time-varying mMNA clinical categories (well-nourished, risk for malnourishment, and malnourished) in order to examine clinically meaningful nutritional associations. In addition, the following covariates were tested in each of the models: dementia duration (from onset age to baseline), age of dementia onset, education, biological sex, dementia type (AD, VaD, vs. other dementia), and presence of the APOE E4 allele. Model fitting began with a base model with the mMNA score/category and time, followed by the addition of covariates sequentially, the order of which was based on exploratory linear mixed effects models that included only the covariates and the outcome. Thus, model building (after the entry of the primary independent variable and a term for time), proceeded with the entry of covariates most highly associated to those not significantly associated with the outcome. Nested models using maximum likelihood (ML) estimation were compared using likelihood ratio (LR) tests to identify the factors that produced the best model fit (p < 0.05). The AIC and BIC were also examined, but when fit indices did not agree, the LR was used as the primary criterion for model fit. Linear and quadratic terms for time were also tested for significance in each model using this method. To obtain parameter estimates, the models were rerun using restricted maximum likelihood (REML). Note that for domain and individual NPI symptom scores, only the continuous mMNA score was examined. Statistical analyses were conducted using SPSS version 28 software.

Results

Characteristics of Study Sample

A total of 328 individuals with dementia were enrolled in the DPS. Thirty-six participants were excluded due to missing data as follows: 35 missing mMNA scores and 1 missing APOE E4 genotype, yielding a total of 292 participants as the final sample. Approximately 56.2% of the sample were women and 71.9% were diagnosed with AD. A significantly higher percentage of those excluded from the final sample resided in an assisted living facility (26.5% vs. 18.9%) or a nursing home or locked unit of an assisted living facility (32.4% vs. 10.3%) but had fewer medical co-morbidities at baseline. Additionally, those excluded from the sample had significantly longer dementia duration from onset age to baseline visit (mean of 4.37 vs. 3.46 years) and lower scores on the Mini-Mental State Exam (mean of 17.15 vs. 20.49 points). Table 1 presents descriptive statistics for those included vs. excluded from the final sample.

Table 1.

Participant Characteristics

Included in
Sample		 Excluded from
Sample		 p-value
Sex at birth .066
  Males N (%) 128 (43.8) 10 (27.8)
  Females N (%) 164 (56.2) 26 (72.2)
Education, M (SD) 13.33 (2.97) 13.34 (2.61) .979
Dementia type (%) .969
  AD 210 (71.9) 26 (72.2)
  Other Dementia 82 (28.1) 10 (27.8)
Age of dementia onset, M (SD) 82.16 (5.96) 84.22 (6.04) .051
Place of Residence at Baseline <.001
  Home 206 (70.8) 14 (41.2)
  Assisted living 55 (18.9) 9 (26.5)
  Locked residential/nursing home 30 (10.3) 11 (32.4)
Number of Medical Conditions at Baseline .001
  M (SD) 4.18 (1.68) 3.22 (1.61)
  Minimum, Maximum 0, 10 1, 8
Number of Medical Conditions at Baseline (Groups) .040
  0-3 111 (38) 21 (58.3)
  4 or 5 119 (40.8) 12 (33.3)
  6 or more 62 (21.2) 3 (8.3)
Baseline dementia duration, .006
  M (SD) 3.46 (1.89) 4.37 (1.84)
  Minimum, Maximum .74, 11.08 1.28, 7.78
Follow-up time in years in overall study (M, SD) 2.53 (2.15) 0.93 <.001
  Minimum, Maximum 0.00, 9.471 (1.15) 0.00, 4.23
Clinical Dementia Rating2 at Baseline, M (SD) 1.34 (0.63) 1.53 (0.75) .170
Mini-Mental State Exam at Baseline3, M (SD) 20.49 (6.87) 17.15 (7.95) .012
NPI-12 total score at Baseline, M (SD) 11.69 (11.21) 12.93 (9.11) .559
NPI domain scores4 at Baseline, M (SD)
  Agitation/aggression 0.78 (1.81) 0.63 (1.36) .638
  Depression 1.39 (2.19) 1.59 (2.38) .625
  Apathy 2.22 (3.33) 2.31 (3.02) .880
  Anxiety 0.92 (1.82) 1.35 (1.85) .208
  Disinhibition 0.39 (1.06) 0.31 (0.86) .699
  Irritability 0.96 (1.84) 0.84 (1.55) .742
  Aberrant motor behavior 0.77 (1.95) 0.91 (1.65) .711
  Nighttime behavior 1.68 (2.94) 1.07 (2.52) .219
  Psychosis 1.39 (3.02) 2.03 (3.37) .259
mMNA total score at Baseline5, M (SD) 16.54 (2.96)
mMNA clinical groups at Baseline5, N (%)
  Malnourished 34 (13.2) 0
  At-risk for malnutrition 94 (36.6) 0
  Well-nourished 129 (50.2) 1
Open in a new tab

Note. M= mean; SD= standard deviation; AD= Alzheimer’s dementia; NPI= Neuropsychiatric Inventory

1

Differs from the maximum of 6.5 years of visits included in analyses due to sparse numbers for follow-ups beyond 6 years

2

Clincal Dementia Rating: .5 = Questionable Dementia, 1 = Mild Dementia, 2 = Moderate Dementia, ≥ 3 = Severe Dementia

3

MMSE maximum possible score = 30.

4

Maximum domain score=12 points with the exception of psychosis (24 points)

5

Differences between mMNA scores/groups were not examined due to missing mMNA for 35 of the 36 excluded from the sample.

Nutritional Status and Neuropsychiatric Symptoms

Total NPI increased over time, by approximately one point per year (b’s = 0.89 and 1.05). Higher nutritional status was associated with lower NPI total scores [b (95% CI) = −0.58 (−0.86, −0.29)] controlling for age of dementia onset, sex, and number of medical conditions. With respect to the clinical mMNA categories, those at risk for malnourishment had NPI total scores nearly two points higher [b (95% CI) =1.76 (.04, 3.48)] while those who were malnourished, approximately three points higher [(b = (95% CI) = 3.20 (0.62, 5.78)] compared to well-nourished participants in models controlling for sex and number of medical conditions. Effect sizes analogous to Cohen’s d standardized mean differences as applied to linear mixed models (Westfall et al., 2014) were 0.31 for those at risk for malnourishment and 0.17 for the malnourished compared to the well-nourished. See Table 2 for the results of the adjusted models for the mMNA and clinical mMNA categories (bottom) and total NPI.

Table 2.

Fully Adjusted Linear Mixed Models of mMNA and Neuropsychiatric Symptoms

Fully adjusted models of mMNA Continuous Score
NPI Total Psychosis Depression Apathy Anxiety Irritability Aberrant
Motor
Behavior
Intercept 35.20*** [18.66, 51.73] 3.47*** [1.63, 5.31] 3.42*** [2.37, 4.47] 5.08*** [3.52, 6.65] 1.25** [0.46, 2.04] 1.24** [0.23, 2.25] 0.91 [−0.11, 1.93]
Time 0.89*** [0.37, 1.41] 0.17 [−0.01, 0.35] −0.08 [−0.18, 0.02] 0.28** [0.11, 0.45] 0.02 [−0.07, 0.12] −0.01 [−0.09, 0.08] 0.16** [0.04, 0.28]
mMNA total score −0.58*** [−0.86, −0.29] −0.08* [−0.16, −0.004] −0.11*** [−0.16, −0.05] −0.19*** [−0.28, −0.11] −0.01 [−0.06, 0.04] 0.03 [−0.02, 0.07] −0.02 [−0.08, 0.03]
Male sex −2.35* [−4.44, −0.25] NA NA NA −0.36* [−0.72, −0.005] NA −0.43* [−0.83, −0.03]
APOE E4 positive NA 0.75** [0.21, 1.28] NA NA NA NA NA
Dementia type NA NA NA NA NA NA
  Any AD −0.46 [−0.25, 0.34]
  Other −1.40*[−2.42, −0.38]
Dementia onset age −0.18* [−0.37, −0.001] NA NA NA NA NA NA
Dementia duration NA 0.17** [0.03, 0.31] −0.11* [−0.21, −0.004] NA NA NA 0.12* [0.02, 0.23]
Education NA −.09* [−0.18, −0.002] NA NA NA −0.07* [−0.12, 0.01] NA
Number of medical conditions NA NA NA NA
  0 to 3 0.93 [−1.12, 2.97] 0.11 [−0.30, 0.51] 0.83* [.17, 1.49] NA 0.01 [−0.32, 0.35]
  6 or more 2.94** [0.74, 5.13] .57* [0.13, 1.02] 0.57 [−0.15, 1.30] NA .58**1 [.21, .95]
Fully adjusted models of mMNA Clinical Categories
NPI Total Psychosis Depression Apathy Anxiety Irritability Aberrant
Motor
Behaviors
Intercept 9.63*** [7.54, 11.72] NA NA NA NA NA NA
Time 1.05*** [0.52, 1.57] NA NA NA NA NA NA
mMNA malnourished 3.20* [0.62, 5.78] NA NA NA NA NA NA
mMNA at risk for malnutrition 1.76* [0.04, 3.48] NA NA NA NA NA NA
Male sex −2.54* [−4.67, −.041] NA NA NA NA NA NA
Number of medical conditions
  0 to 3 .61 [−1.45, 2.67] NA NA NA NA NA NA
  6 or more 3.00** [0.78, 5.22] NA NA NA NA NA NA
Open in a new tab

Note. Parameter estimates and 95% confidence intervals (presented in brackets) presented in the table are based on linear mixed effects models using restricted maximum likelihood. mMNA= modified Mini Nutritional Assessment; APOE= Apolipoprotein; Reference category for number of medical conditions is 4 to 5 conditions. Column headings represent individual outcomes.

1

Pairwise comparisons also revealed significantly higher irritability score for 6 or more medical conditions compared to 0-3 (by 0.56, p = .007; not displayed in the table). NA= not applicable as the variable was not included in the model.

*

p < .05

**

p < .01

***

p < .001

In the analysis of NPI domains, higher mMNA total score was associated with lower scores on psychosis [b (95% CI) = −0.08 (−0.16, −0.004)], depression [b (95% CI = −0.11 (−0.16, −0.05], and apathy [b (95% CI = −0.19 (−0.28, −0.11)]. See Table 2 for the adjusted models for each NPI domain.

Amongst the covariates of the various models, men had significantly lower NPI total scores (by approximately 2 to 2.50 points) and domain scores for anxiety and aberrant motor behaviors than women (both by approximately 0.4 points). The APOE E4 allele was associated with a higher psychosis score by 0.75 points. Type of dementia was significantly associated with psychosis in that those with VaD scored approximately 1.4 points lower than those with other dementias, and longer dementia duration at baseline was associated with slightly lower depression scores, but slightly higher psychosis and aberrant motor behavior. Years of educational attainment were associated with slightly lower psychosis and anxiety. Number of medical conditions were associated with total NPI score and domains of apathy, depression, and irritability. Those with 6+ conditions had significantly higher total NPI (by approximately 3 points) and depression and irritability (all by approximately 0.5 points) compared to those with 4-5 conditions. Those with 0-3 conditions also had higher apathy (by 0.83 points) than individuals with 4-5 conditions, and lower irritability (by 0.56 points) than individuals with 6+ conditions.

Domain scores of anxiety, irritability, and aberrant motor behavior were not significantly associated with mMNA total score, and the models for agitation/aggression, night-time behaviors and disinhibition failed to converge.

Discussion

Worse nutritional status was associated with higher NPS over time, in those with dementia. Compared to well-nourished participants, those in the malnourished category were rated over three points higher on the total NPI and those at risk for malnutrition, nearly two points higher. Specific symptom domains affected by poorer nutritional status were psychosis, depression, and apathy. These results are consistent with previous findings from the REAL French study that reported higher NPI total scores in individuals with probable AD (Brocker et al., 2003; Guerin et al., 2005) and a cross-sectional Italian study reporting that apathy and hallucinations were associated with risk for malnutrition (Spaccavento et al., 2009). Similarly, Roque et al. (2013) reported significant associations between hallucinations and malnutrition, among other NPI domains in their cross-sectional analysis. Apathy was also associated with nutritional status in a crosssectional, clinic-based study of Japanese women with MCI or early AD (Kimura et al., 2019).

Among other NPS, aberrant or aggressive motor behavior (Roque et al., 2013; Kishino et al., 2022), behavioral disturbance (wandering, pacing; Kimura et al., 2019), agitation, anxiety, and sleep behaviors (Roque et al., 2013) have also been associated with malnutrition. We did not find similar associations in our sample and statistical modeling of agitation failed convergence. The present findings of NPS add to our prior work, that poorer nutritional status is associated with worse cognitive and functional status (Sanders et al., 2016) and risk for severe dementia and mortality (Sanders et al., 2018).

We also found men to have lower total NPI scores as well as aberrant motor behaviors and anxiety than women. Prior studies have reported men to experience less severe symptoms of psychosis (Spalletta et al., 2010; Eikelboom et al., 2021), aberrant motor behaviors (Eikelboom et al., 2021), anxiety (Spalletta et al., 2010), depression (Eikelboom et al., 2021), and lower total NPI (Spalletta et al., 2010; Tao et al., 2018). Eikelboom et al. (2021) also reported men to have higher levels of apathy but not NPI total in that meta analysis. . While lower levels of education have previously been reported with elevated agitation and irritability (Gabryelewicz et al., 2002; Apostolova et al., 2014), we found an association with irritability and psychosis. Medical comorbidities were associated with overall severity of NPS and several individual symptoms, similar in effect sizes to clinical malnutrition categories (results not shown).

The significance of our results in clinical mMNA categories suggests relatively small effects with total NPI scores, suggesting a 3-point difference amongst those meeting criteria for malnourishment and even smaller effects for domain scores. However, as reported in caregiver studies, caregiver burden may vary by even slight increases in total NPI (Kawano et al., 2020) and by specific or individual NPS (Terum et al., 2017). Furthermore, as the occurrence of NPS is likely multi-determined, identifying and addressing one of a number of possible contributing factors may prove beneficial.

NPS are difficult to treat (Lyketsos et al., 2011) and non-pharmacological strategies are considered preferred, “first-line” approaches (Kales et al., 2019). Nutritional status declines in older adults (Starr et al., 2015) and more prominently in persons with dementia (Cortes et al., 2008). Although causal relationships cannot be drawn from our study, addressing nutritional deficiencies in dementia may be beneficial. A small number of nutrient-based, randomized controlled trials have been conducted in ADRD and have found a positive effect on cognitive status (see meta-analysis by Allen et al, 2013). However, a recent review reported mixed effects of supplements on cognitive and functional decline the results of which varied by severity of dementia (see Vlachos & Scarmeas, 2019). Notably, none of the studies included in either metaanalysis/review examined effects on NPS (Allen et al., 2013; Vlachos & Scarmeas, 2019). In view of our current findings, trials of dietary interventions may benefit those vulnerable or at risk for malnutrition.

Greenwood et al. (2005) posed that rather than cognitive status, it is the behavioral deterioration of eating habits (e.g. choosing less protein-rich foods) that may be associated with some of the observed NPS in individuals with AD. Consistent with this notion, prior work in the DPS found that varied protein sources (a subcomponent of the MNA score) may be beneficial with respect to risk for severe dementia (Sanders et al., 2016), though this has not been examined for NPS. Additionally, a nutritional intervention in Spain aimed to educate physicians and caregivers on dietary recommendations (e.g., promoting a balanced diet), eating behaviors (e.g., substituting foods that are rejected), and dietary modifications, found a reduction in risk for malnutrition in those with mild-to-moderate dementia (Salva et al., 2009). Thus, nutritional education and monitoring to prevent a state of malnutrition as part of behavioral and lifestyle interventions (Gitlin et al., 2012; Kales et al., 2014) may be beneficial to reducing NPS. Although causal relationships cannot be made from our and others’ observational studies, routine monitoring of nutrient intake and nutritional status for persons with dementia is recommended as standard of care given their associations with cognitive, functional and NPS outcomes.

Our study had several strengths and expanded upon previous studies by its use of a population-based sample, high participation rates and extended follow-up (up to six years) of individuals identified relatively early in their course of dementia. This study also examined individual NPS as well as an extensive number of medical comorbidities which along with nutritional status, were ascertained at each visit and incorporated as time-varying variables in analyses. Nonetheless, several study limitations are noted, including insufficient numbers to thoroughly assess differences by underlying dementia type. Due to the homogeneous nature of this sample, we were unable to assess other potential confounding factors such as race/ethnicity and socioeconomic status (SES). SES, which is an important factor that may contribute to food insecurity, was not available in this dataset. Our findings may also not generalize to those living in institutions as the majority of those included in the sample were residing in private homes. Additionally, many statistical tests were run in our examination of total NPI and domain scores, raising the potential of a Type 1 error; thus, results should be replicated in other samples.

In conclusion, we found nutritional status, malnourishment and medical comorbidities were significantly associated with NPS and specific symptom domains in dementia. Together with our prior work on cognitive and functional outcomes, our results suggest that the prevention of malnourishment may be an important goal in the care for persons with dementia.

Acknowledgements:

Funded by grants from the National Institute on Aging R01AG11380, R01AG21136, R01AG18712

Footnotes

Conflicts of Interest - Dr. Lyketsos receives funding Roche, Avanir, Karuna, Maplight, Axsome, GIA, GW Research Limited, Orion, Otsuka, Lundbeck, Merck, EXCIVA GmbH.

References

  1. Allen VJ, Methven L, Gosney MA (2013). Use of nutritional complete supplements in older adults with dementia: Systematic review and meta-analysis of clinical outcomes. Clinical Nutrition. 32(6), 950–957. doi: 10.1016/j.clnu.2013.03.015 [DOI] [PubMed] [Google Scholar]
  2. American Psychiatric Association. (1987). Diagnostic and Statistical Manual of Mental Disorders (3rd ed). Revised. [Google Scholar]
  3. Apostolova LG, Di LJ, Duffy EL, et al. (2014). Risk Factors for Behavioral Abnormalities in Mild Cognitive Impairment and Mild Alzheimer’s Disease. Dement Geriatr Cogn Disord, 37(5–6):315–326. doi: 10.1159/000351009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bjoerke-Bertheussen J, Ehrt U, Rongve A, Ballard C, Aarsland D. (1987). Neuropsychiatric symptoms in mild dementia with Lewy bodies and Alzheimer’s disease. Dementia and geriatric cognitive disorders, 34(1):1–6. [DOI] [PubMed] [Google Scholar]
  5. Breitner JCS, Wyse BW, Anthony JC, et al. (1999). APOE- 4 count predicts age when prevalence of AD increases, then declines: The Cache County Study. Neurology, 53(2):321–321. doi: 10.1212/WNL.53.2.321. [DOI] [PubMed] [Google Scholar]
  6. Brocker P, Benhamidat T, Benoit M, StaccinP P, Bertogliati C, Gurrin O. (2003). Etat nutritionnel et maladie d’Alzheimer • rrsultats prrliminaires de l’rtude REAL.FR. :5. [DOI] [PubMed] [Google Scholar]
  7. Cortes F, Nourhashémi F, Guérin O, et al. (2008). Prognosis of Alzheimer’s disease today: A two-year prospective study in 686 patients from the REAL-FR Study. Alzheimer’s & Dementia, 4(1):22–29. doi: 10.1016/j.jalz.2007.10.018. [DOI] [PubMed] [Google Scholar]
  8. Cummings JL, Mega M, Gray K, Rosenberg-Thompson S, Carusi DA, Gornbein J. (1994). The Neuropsychiatric Inventory: Comprehensive assessment of psychopathology in dementia. Neurology, 44(12):2308–2308. doi: 10.1212/WNL.44.12.2308. [DOI] [PubMed] [Google Scholar]
  9. Cummings JL. (1997). The Neuropsychiatric Inventory: Assessing psychopathology in dementia patients. Neurology, 48(Issue 5, Supplement 6):10S–16S. doi: 10.1212/WNL.48.5Suppl_6.10S. [DOI] [PubMed] [Google Scholar]
  10. Delano-Wood L, Houston WS, Emond JA, et al. (2008). APOE genotype predicts depression in women with Alzheimer’s disease: a retrospective study. International Journal of Geriatric Psychiatry: A journal of the psychiatry of late life and allied sciences, 23(6):632–636. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Dorner B. (2010). Position of the American Dietetic Association: Individualized Nutrition Approaches for Older Adults in Health Care Communities. Journal of the American Dietetic Association, 110(10):1549–1553. doi: 10.1016/j.jada.2010.08.022. [DOI] [PubMed] [Google Scholar]
  12. Eikelboom WS, Pan M, Ossenkoppele R, et al. (2021). Sex differences in neuropsychiatric symptoms in Alzheimer’s disease dementia: A meta-analysis. Alzheimer’s & Dementia, 17:e055542. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Fernández-Martínez M, Castro J, Molano A, Zarranz JJ, Rodrigo RM, Ortega R. (2008). Prevalence of neuropsychiatric symptoms in Alzheimer’s disease and vascular dementia. Current Alzheimer Research, 5(1):61–69. [DOI] [PubMed] [Google Scholar]
  14. Gabryelewicz T, Religa D, Styczynska M, et al. (2002). Behavioural pathology in Alzheimer’s disease with special reference to apolipoprotein E genotype. Dementia and Geriatric Cognitive Disorders, 14(4):208–212. [DOI] [PubMed] [Google Scholar]
  15. Gitlin LN, Kales HC, Lyketsos CG. (2012). Nonpharmacologic management of behavioral symptoms in dementia. Jama, 308(19):2020–2029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Greenwood CE, Tam C, Chan M, Young KW, Binns MA, Van Reekum R. (2005). Behavioral disturbances, not cognitive deterioration, are associated with altered food selection in seniors with Alzheimer’s disease. The Journals of Gerontology Series A: Biological Sciences and Medical Sciences, 60(4):499–505. [DOI] [PubMed] [Google Scholar]
  17. Guerin O, Soto ME, Brocker P, Robert PH, Benoit M, Vellas B. (2005). Nutritional status assessment during Alzheimer’s disease: results after one year (the REAL French Study Group). J Nutr Health Aging, 9(2):81–84. [PubMed] [Google Scholar]
  18. Guigoz Y, Vellas B, Garry PJ. (1997). Mini nutritional assessment: a practical assessment tool for grading the nutritional state of elderly patients. The mini nutritional assessment: MNA Nutrition in the elderly,15–60. [DOI] [PubMed] [Google Scholar]
  19. Kales HC, Gitlin LN, Lyketsos CG. (2014). Assessment DEP on the, Dementia M of the NS of. Management of neuropsychiatric symptoms of dementia in clinical settings: recommendations from a multidisciplinary expert panel. Journal of the American Geriatrics Society, 62(4):762–769. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kales HC, Lyketsos CG, Miller EM, Ballard C. (2019). Management of behavioral and psychological symptoms in people with Alzheimer’s disease: an international Delphi consensus. International Psychogeriatrics, 31(1):83–90. doi: 10.1017/S1041610218000534. [DOI] [PubMed] [Google Scholar]
  21. Kawano Y, Terada S, Takenoshita S, et al. (2020). Patient affect and caregiver burden in dementia. Psychogeriatrics, 20(2):189–195. doi: 10.1111/psyg.12487. [DOI] [PubMed] [Google Scholar]
  22. Kimura A, Sugimoto T, Kitamori K, et al. (2019). Malnutrition is associated with behavioral and psychiatric symptoms of dementia in older women with mild cognitive impairment and early-stage Alzheimer’s disease. Nutrients, 11(8):1951. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kishino Y, Sugimoto T, Kimura A, et al. (2022). Longitudinal association between nutritional status and behavioral and psychological symptoms of dementia in older women with mild cognitive impairment and early-stage Alzheimer’s disease. Clinical Nutrition, 41(9):1906–1912. [DOI] [PubMed] [Google Scholar]
  24. Kwak S, Park S, Kim J, Park S, Lee JY. (2020). Multivariate neuroanatomical correlates of behavioral and psychological symptoms in dementia and the moderating role of education. NeuroImage: Clinical, 28:102452. doi: 10.1016/j.nicl.2020.102452. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lyketsos CG, Carrillo MC, Ryan JM, et al. (2011). Neuropsychiatric symptoms in Alzheimer’s disease. Alzheimer’s & Dementia, 7(5):532–539. doi: 1016/j.jalz.2011.05.2410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lyketsos CG, Galik E, Steele C, Steinberg M, Rosenblatt A, Warren A, Sheppard JM, Baker A, Brandt J (1999). The General Medical Health Rating: a bedside global rating of medical comorbidity in patients with dementia. Journal of the American Geriatric Society, 47, 487–91. [DOI] [PubMed] [Google Scholar]
  27. McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM. (1984). Clinical diagnosis of Alzheimer’s disease: Report of the NINCDS-ADRDA Work Group* under the auspices of Department of Health and Human Services Task Force on Alzheimer’s Disease. Neurology, 34(7):939–939. doi: 10.1212/WNL.34.7.939 [DOI] [PubMed] [Google Scholar]
  28. Miech RA, Breitner JCS, Zandi PP, Khachaturian AS, Anthony JC, Mayer L. (2002). Incidence of AD may decline in the early 90s for men, later for women: The Cache County study. Neurology, 58(2):209–218. [DOI] [PubMed] [Google Scholar]
  29. Norman K, Haß U, Pirlich M. (2021). Malnutrition in Older Adults—Recent Advances and Remaining Challenges. Nutrients, 13(8):2764. doi: 10.3390/nu13082764 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Román GC, Tatemichi TK, Erkinjuntti T, et al. (1993). Vascular dementia: diagnostic criteria for research studies: report of the NINDS-AIREN International Workshop. Neurology, 43(2):250–250. [DOI] [PubMed] [Google Scholar]
  31. Roque M, Salva A, Vellas B. (2013). Malnutrition in community-dwelling adults with dementia (NutriAlz Trial). The journal of nutrition, health & aging, 17(4):295–299. [DOI] [PubMed] [Google Scholar]
  32. Salva A, Andrieu S, Fernandez E, et al. (2009). Health and nutritional promotion program for patients with dementia (NutriAlz Study): design and baseline data. JNHA-The Journal of Nutrition, Health and Aging, 13(6):529–537. [DOI] [PubMed] [Google Scholar]
  33. Sanders C, Behrens S, Schwartz S, et al. (2016). Nutritional status is associated with faster cognitive decline and worse functional impairment in the progression of dementia: the cache county dementia progression study. Journal of Alzheimer’s disease, 52(1):33–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Sanders CL, Wengreen HJ, Schwartz S, et al. (2018). Nutritional Status is Associated with Severe Dementia and Mortality: The Cache County Dementia Progression Study. Alzheimer Dis Assoc Disord, 32(4):298–304. doi: 10.1097/WAD.0000000000000274. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Spaccavento S, Del Prete M, Craca A, Fiore P. (2009). Influence of nutritional status on cognitive, functional and neuropsychiatric deficits in Alzheimer’s disease. Archives of Gerontology and Geriatrics, 48(3):356–360. doi: 10.1016/j.archger.2008.03.002. [DOI] [PubMed] [Google Scholar]
  36. Spalletta G, Musicco M, Padovani A, et al. (2010). Neuropsychiatric symptoms and syndromes in a large cohort of newly diagnosed untreated patients with Alzheimer disease. The American Journal of Geriatric Psychiatry, 18(11):1026–1035. [DOI] [PubMed] [Google Scholar]
  37. Starr KNP, McDonald SR, Bales CW. (2015). Nutritional vulnerability in older adults: a continuum of concerns. Current nutrition reports, 4(2):176–184. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Steinberg M, Shao H, Zandi P, et al. (2008). Point and 5-year period prevalence of neuropsychiatric symptoms in dementia: the Cache County Study. International Journal of Geriatric Psychiatry, 23(2):170–177. doi: 10.1002/gps.1858. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Tao Y, Peters ME, Drye LT, et al. (2018). Sex differences in the neuropsychiatric symptoms of patients with Alzheimer’s disease. American Journal of Alzheimer’s Disease & Other Dementias®, 33(7):450–457. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Terum TM, Andersen JR, Rongve A, Aarsland D, Svendsboe EJ, Testad I. (2017). The relationship of specific items on the Neuropsychiatric Inventory to caregiver burden in dementia: a systematic review. International Journal of Geriatric Psychiatry, 32(7):703–717. doi: 10.1002/gps.4704. [DOI] [PubMed] [Google Scholar]
  41. Tschanz JT, Corcoran CD, Schwartz S, et al. (2011). Progression of cognitive, functional, and neuropsychiatric symptom domains in a population cohort with Alzheimer dementia: the Cache County Dementia Progression study. The American Journal of Geriatric Psychiatry, 19(6):532–542. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Tschanz JT, Welsh-Bohmer KA, Plassman BL, et al. (2002). An adaptation of the modified mini-mental state examination: analysis of demographic influences and normative data: the cache county study. Cognitive and Behavioral Neurology, 15(1):28–38. [PubMed] [Google Scholar]
  43. Vellas B Guigoz Y, Garry PJ, Nourhashemi F, Bennahum D, Lauque S, Albarede JL (1999). The Mini Nutritional Assessment (MNA) and its use in grading the nutritional state of elderly patients. Nutrition, 15, 116–22. [DOI] [PubMed] [Google Scholar]
  44. Vernon EK, Cooley B, Rozum W, et al. (2019). Caregiver-care recipient relationship closeness is associated with neuropsychiatric symptoms in dementia. The American Journal of Geriatric Psychiatry, 27(4):349–359. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Vlachos GS, & Scarmeas N (2019). Dietary interventions in mild cognitive impairment and dementia. Dialogues in Clinical Neuroscience, 21(1): 69–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Westfall J, Kenny DA, & Judd CM (2014). Statistical power and optimal design in experiments in which samples of participants respond to samples of stimuli. Journal of Experimental Psychology: General, 143(5): 2020–2045. [DOI] [PubMed] [Google Scholar]

RESOURCES