Delirium Severity and Physical Function in Hospitalized Persons Living with Dementia: Moderation by Age, Sex, and Race

Abstract

Background:

This study investigated whether demographic characteristics (age, sex, and race) moderated delirium severity as a predictor of physical function in hospitalized persons living with dementia.

Methods:

The sample consisted of 351 patients enrolled in a randomized controlled trial (FFC-AC-EIT). Preliminary analysis was conducted to assess the main effect, and multiple linear regression was used to examine the moderating effect of demographic characteristics between delirium severity and physical function.

Results:

Both age and sex were found to have significant moderating effects on the relationship between delirium severity and physical function (β = 2.22; p = 0.02 and β = 1.34; p = 0.04, respectively). Older adults aged ≥ 85 years with higher levels of delirium severity reported lower levels of physical function compared to older adults aged 65–84 years. Males with higher levels of delirium severity reported lower levels of physical function compared to females. Race did not significantly moderate the association between delirium severity and physical function (β = 0.22; p = 0.90).

Conclusions:

Our findings suggest that age and sex may have differential effects on physical function across different levels of delirium severity in hospitalized persons living with dementia.

Introduction

Delirium is a common occurrence among hospitalized older adults living with dementia, and it poses significant risks to their health and wellbeing.^1,2 Delirium is characterized by sudden changes in mental status, including confusion, disorientation, and impaired attention.² Hospitalized persons living with dementia have the highest risk for delirium,^1,3,4 with a prevalence rate of up to 89% during hospitalization.⁵ When compared to patients without dementia, those with dementia are at a 3 to 4 times higher risk of developing delirium during their hospital stay.⁶ Delirium is also associated with a significant mortality risk; between 24% and 76% of patients who experience a delirium episode die within one year.⁷ The unfamiliar hospital environment, disruption of normal routines, the acute illness, and non-dementia friendly medications and care practices all contribute to the distress of persons living with dementia,⁸ and contribute to an increased risk of delirium, as well as other adverse events.⁹

The presence of delirium among older adults, especially those with dementia, has consistently been associated with prolonged hospital stays and resulted in adverse effects on physical function.^1,4 In older adults, physical function plays a crucial role in preventing falls, rehospitalizations, infections, long-term care placement, and early mortality.^10–15 Decline in physical function, characterized by a decrease in the ability to perform daily activities, is observed in approximately 30–60% of adults aged 65 and older during their hospitalization.^11,16,17 This decline is often attributed to extended periods of inactivity while hospitalized.^11,16,17 In addition to delirium, the risk of functional loss during hospitalization is heightened by other clinical factors, such as baseline and admission function, pain, cognition, polypharmacy, depression, length of hospital stay, restraints and other tethers (e.g., pulse oximetry, catheters, intravenous), and comorbidities.^16,18–22

Increased risk of delirium during hospitalization has been linked to factors such as advanced age,² male sex,² and Black race.²³ Similarly, older age,²¹ male sex,²¹ and identifying as Black²³ have been identified as predictors of functional decline in hospitalized older adults. However, these studies did not examine the interaction effects of age, sex, and race on delirium severity and their impact on physical function in hospitalized persons living with dementia. Thus, little is known about the moderating role of these demographic characteristics underlying the relationship between delirium and physical function.

To address this research gap, the aim of this study was to investigate whether demographic characteristics (i.e., age, sex, and race) moderate any differences in delirium severity as predictors of physical function in hospitalized persons living with dementia. Based on prior research,^2,21,23 we hypothesized that sex, age, and race would interact with delirium severity in predicting physical function. Specifically, we predicted that Black, male, and older patients with higher levels of delirium would have lower levels of physical function. Understanding the influence of these demographic characteristics across various levels of delirium can help inform the design and focus of future interventions to optimize physical function in hospitalized persons living with dementia.

Methods

Design

This descriptive study utilized baseline data from an ongoing research study called Function Focused Care for Acute Care Using the Evidence Integration Triangle (FFC-AC-EIT), which is registered in ClinicalTrials.gov under the identifier NCT04235374. The primary goal of FFC-AC-EIT was to assist hospital staff in actively involving patients with dementia in physical activity during care interactions, with the aim of enhancing the overall physical function of their patients. The study received approval from the Institutional Review Board at the University of Maryland, Baltimore and has been published.²⁴ All participants provided written informed consent prior to data collection.

Sample

This study included patients from ten hospitals (one medical unit per hospital) in Maryland and Pennsylvania. Patients were eligible to participate if they were 65 years of age or older, admitted to a medical unit with a medical diagnosis (excluding patients with COVID-19), and screened positive for dementia on various assessments. These assessments included: (1) a score of ≤ 20 on the Saint Louis University Mental Status (SLUMS) exam, while those with less than a high education, the cutoff was a score of ≤ 19²⁵; (2) a score of ≥ 2 on the AD8 Dementia Screening interview²⁶; (3) a Clinical Dementia Rating Scale (CDR) score ranging from 0.5 to 2.0²⁷; and (4) a Functional Activities Questionnaire (FAQ) score of ≥ 9.²⁸ Patients were excluded from the study if they were enrolled in hospice, had been on the unit for more than 48 hours, lacked a family member or caregiver to contact, were expected to undergo surgery, or had a major acute psychiatric disorder or a significant neurological condition affecting cognition other than dementia.

Procedures

Trained research evaluators who had prior experience working with this population collected data in various methods including chart review, patient observation, and report from hospital staff. The evaluators were blinded to the intervention. Data collection took place within 48 hours of the patients’ admission to the unit, prior to the implementation of the FFC-AC-EIT intervention.

Measures

The independent variable was delirium severity, assessed using the Confusion Assessment Method Short Form (CAM-S).²⁹ The CAM-S scale comprises four items that assess acute onset, inattention, disorganized thinking, and altered level of consciousness as indicators of delirium. Each item is scored on a scale ranging from 0–7, with higher scores reflecting more severe delirium. Previous research on the CAM-S has shown robust psychometric properties, supporting its reliability and validity.^29,30

The dependent variable was physical function, evaluated using the Barthel Index (BI), which is a ten-item instrument.³¹ The BI measures an individual’s performance in activities of daily living, specifically related to mobility, self-care, and bowel/bladder functions. The total scores on the BI range from 0, indicating complete dependence, to 100, representing total independence. Acceptable validity and reliability of the BI have been established.^31,32

The moderator variables, obtained from chart review, included three demographic characteristics: sex (classified as male or female), age (categorized as 65–84 and ≥ 85), and race (Black/White).

The covariates included education, marital status, cognition, comorbidities, and pain. These factors have been previously associated with delirium severity or physical function.^{2,16,18,20,21,33,34} Education, marital status, and comorbidities were gathered through chart review, with comorbidities assessed using the Charlson Comorbidity Index (CCI).³⁵ Cognition was evaluated using SLUMS, which measures various aspects of memory, including orientation, problem-solving, and recall. Scores on the SLUMS range from 0–30, with lower scores indicating more severe cognitive impairment. Prior research of the SLUMS has demonstrated adequate psychometric properties.³⁶ Pain was assessed using the Pain Assessment in Advanced Dementia (PAINAD).³⁷ The assessment covers five items/domains: breathing, negative vocalizations, facial expression, body language, and consolability. Each item is scored on a scale from 0–2. The total scores on the PAINAD range from 0–48, with specific ranges indicating different levels of pain severity: 1–3 for mild pain, 4–6 for moderate pain, and 7–10 for severe pain. Prior work confirmed internal reliability (α = 0.90) of the PAINAD, as well as support for interrater agreement (κ = .74) and concurrent validity (Kendall’s τ = 0.73).³⁸

Data Analysis

Participant demographic and clinical characteristics were reported using frequencies and percentages for categorical variables and mean ± standard deviation (SD) for continuous variables. Data for missing values, existence of outliers, and normality was checked. All variables satisfied the assumptions of multicollinearity. Before examining moderator effects, preliminary analysis was conducted to assess the main effect between the independent variable (delirium severity) and dependent variable (physical function).

To investigate the moderating effects of demographic characteristics (i.e., age, sex, and race), multiple linear regression was performed with delirium severity as the independent variable and physical function as the dependent variable (see Figure 1). Two-way interaction terms were calculated by multiplying delirium severity by each demographic characteristic (i.e., delirium severity*age, delirium severity*sex, and delirium severity*race). The interaction term must be statistically significant to determine a moderating effect. The continuous predictor variable, delirium severity, was mean centered prior to the moderation analysis. Graphics were plotted for better understanding of any significant interactions. The model included education, marital status, cognition, comorbidities, and pain as covariates. The alpha level was set to 0.05 to determine statistical significance and SPSS Version 27 was utilized for all analyses (IBM Corp, Armonk, NY).

Figure 1.

Moderation Model by Demographic Characteristics

Note: M1, M2, M3 = Moderator

Results

A total of 351 patients participated in this study. As shown in Table 1, most participants were female (62.1%, n = 218), White (74.9%, n = 263), and non-Hispanic/Latino (98.9%, n = 347). In the 65–84 age group, the mean age was 77.48 ± 4.96 years, and in the ≥ 85 age group, the mean age was 90.02 ± 4.01 years. Among the participants, 38.7% (n = 136) were married and 83.5% (n = 293) had high school or higher-level education. On average, the participants had 3.10 ± 2.35 comorbid conditions and moderate to severe cognitive impairment (SLUMS; 8.01 ± 6.03). Additionally, the sample had substantial functional dependency (BI; 51.35 ± 27.12), low pain (PAINAD; 1.03 ± 1.80), and mild delirium severity (CAM-S; 1.93 ± 1.81). The main effect analysis revealed a significant association between delirium severity and physical function (β = −3.38; p < 0.001).

Table 1.

Characteristics of Patients (N=351): Counts (%) for Categorical Variables and Mean ± Standard Deviation (SD) for Continuous Variables

Variable		n (%) or mean± SD
Age Group (years)	65–84	77.48 ± 4.96
	≥85	90.02 ± 4.01
Sex	Female	218 (62.1)
Race	White	263 (74.9)
	Black	88 (25.1)
Ethnicity	Not Hispanic/Latino	347 (98.9)
	Hispanic/Latino	4 (1.2)
Education	<High school	58 (16.5)
	≥High school graduate	293 (83.5)
Marital Status	Not Married	215 (61.3)
	Married	136 (38.7)
Cognition (SLUMS, range: 0–20)		8.01 ± 6.03
Comorbidities (CCI, range: 0–13)		3.10 ± 2.35
Delirium Severity (CAM-S; range 0–7)		1.93 ± 1.81
Physical Function (BI, range 3–100)		51.35 ± 27.12
Pain (PAINAD, range 0–9)		1.03 ± 1.80

Note: SLUMS = Saint Louis University Mental Status Examination; CCI = Charlson Comorbidity Index; CAM-S = Confusion Assessment Method Short Form; BI = Barthel Index; PAINAD; Pain Assessment in Advanced Dementia.

The results from the moderation analysis are displayed in Table 2. Both age and sex were found to have significant moderating effects on the relationship between delirium severity and physical function (β = 2.22; p = 0.02 and β = 1.34; p = 0.04, respectively). As shown in Figure 2, both age groups exhibit a negative slope, indicating that higher delirium severity is associated with lower physical function. Notably, older adults aged ≥ 85 years with higher levels of delirium severity reported lower levels of physical function compared to older adults aged 65–84 years. Likewise, Figure 3 displays a negative slope for both sexes, demonstrating that higher delirium severity is linked to lower physical function. However, males with higher levels of delirium severity reported lower levels of physical function compared to females. The moderation analysis revealed that race did not significantly moderate the association between delirium severity and physical function (β = 0.22; p = 0.90).

Table 2.

Moderation Effects of Demographic Variables on the Relationship between Delirium Severity and Physical Function (N=351): Coefficients, Standard Errors (SE), and P-Values from a Multiple Linear Regression Model Including Interaction Terms for Delirium Severity and Demographic Characteristics (Age, Sex, and Race)

Variable	Physical Function
	B	SE	p
Delirium Severity	−2.14	2.43	0.03
Age (ref: 65–84 years)	−2.97	2.88	0.10
Sex (ref: Female)	−0.79	2.94	0.09
Race (ref: White)	2.87	3.24	0.38
Delirium Severity x Age	2.22	1.50	0.02
Delirium Severity x Sex	1.34	1.61	0.04
Delirium Severity x Race	0.22	1.80	0.90
Controls
Cognition	1.50	0.25	0.001
Comorbidities	−0.84	0.59	0.15
Marital Status (ref: Not Married)	0.15	0.91	0.87
Education (ref: ≥High school graduate)	−1.39	0.88	0.12

Figure 2.

Moderator Effect of Age on the Relationship between Delirium Severity and Physical Function

Note: Delirium severity was mean centered, thus, a negative value indicates values lower than the mean.

Figure 3.

Moderator Effect of Sex on the Relationship between Delirium Severity and Physical Function

Note: Delirium severity was mean centered, thus, a negative value indicates values lower than the mean.

Discussion

The present study explored whether age, sex, and race moderated delirium severity as a predictor of physical function in hospitalized persons living with dementia. The study found that patients with higher delirium severity had lower function, aligning with previous studies conducted in hospital settings.^2,21,23

Findings of the moderation analysis revealed significant interactions between age and delirium severity, as well as sex and delirium severity, suggesting that age and sex play a moderating role on the association between delirium severity and physical function in hospitalized persons living with dementia. Specifically, older adults ≥ 85 years of age with higher delirium severity reported lower physical function compared to those aged 65–84 years. Additionally, males with higher delirium severity reported lower physical function compared to females. On the other hand, race did not moderate the association between delirium severity and physical function. Our results provided partial support for our hypothesis by indicating that both age and sex had a moderating effect on the relationship between delirium severity and physical function.

The results are consistent with previous studies where age and sex (i.e., older age and male sex) have direct effects on either delirium or physical function in hospitalized older adults with or without cognitive impairment^2,21. Our study adds to literature by examining the interplay between age, sex, and race in relation to delirium severity and its impact on physical function among hospitalized persons living with dementia. The recognition of age and sex differences in the relationship between delirium and physical function has the potential to inform the development of more effective strategies for the assessment, intervention, and prevention of delirium and functional decline in hospitalized persons with dementia. Further investigations should explore the longitudinal relationships between delirium severity, demographic characteristics (i.e., age, sex, and race), and physical function among hospitalized persons living with dementia.

Despite previous research suggesting a direct association between race and either delirium or physical function,²³ our study did not find that race moderates the relationship between delirium severity and physical function. It is important to interpret our findings cautiously, as the majority of our sample exhibited mild delirium and moderate physical function. Therefore, it is possible that race may moderate the main effect of delirium and function in samples that have higher levels of both delirium and lower function in hospitalized persons living with dementia. In addition, this study did not assess and consequently did not control for other potential factors such as medications,^2,22 depression,^20,22 and social support²¹ that may influence the moderating role of race on the relationship between delirium and physical function among hospitalized older adults. Subsequent investigations should explore the extent to which these factors can influence the moderating role of age, sex, and race on the association between delirium severity and physical function.

Strengths and Limitations

Strengths of this study include a substantial sample size, specifically focusing on a population that is often underrepresented: hospitalized persons living with dementia. Additionally, the study explored the interplay between delirium severity, demographic characteristics, and physical function, a critical area that has received inadequate attention within this vulnerable population.

This study has limitations. First, the cross-sectional design restricts our ability to establish causality. Additionally, the generalizability of our findings was limited since the study only included patients from ten hospitals in two states. Moreover, the reliance on self-report data and observations during data collection may introduce biases related to social desirability and memory issues. To address these limitations, future research should consider incorporating patient/caregiver perceptions of the experience of delirium and objective measures (i.e., actigraphy or physical activity).

Conclusions

Our findings suggest that age and sex may have differential effects on physical function across different levels of delirium severity in hospitalized persons living with dementia. The identification of age and sex as moderators is important for the design of delirium guidelines to enhance physical function in the hospital setting.³⁹ By considering the differential effects of age and sex on the relationship between delirium severity and physical function, future clinical interventions can be tailored to address the unique needs and challenges faced by specific age groups and sexes in this population. Further research is warranted to explore the moderating role of demographic characteristics between delirium and physical function in persons with diverse medical acuity and diagnoses across various geographic settings.