Impact of Nutritional Supplementation and a Psychomotor Program on Patients With Alzheimer’s Disease

Abstract

This study aims to evaluate the impact of oral nutritional supplementation (ONS) and a psychomotor rehabilitation program on nutritional and functional status of community-dwelling patients with Alzheimer’s disease (AD). A 21-day prospective randomized controlled trial was conducted and third intervention group performed a psychomotor rehabilitation program. Patients were followed up for 180 days. Mean (standard deviation) score of Mini Nutritional Assessment (MNA) increased both in the nutritional supplementation group (NSG; n = 25), 0.4 (0.8), and in the nutritional supplementation psychomotor rehabilitation program group (NSPRG; n = 11), 1.5 (1.0), versus −0.1 (1.1) in the control group (CG; n = 43), P < .05. Further improvements at 90-day follow-up for MNA in NSG: 1.3 (1.2) and NSPRG: 1.6 (1.0) versus 0.3 (1.7) in CG (P < .05) were observed. General linear model analysis showed that the NSG and NSPRG ▵MNA score improved after intervention, at 21 days and 90 days, was independent of the MNA and Mini-Mental State Examination scores at baseline (Ps > .05). The ONS and a psychomotor rehabilitation program have a positive impact on long-term nutritional and functional status of patients with AD.

Introduction

Although nutritional deficiencies are very common in older adults, especially among patients with probable Alzheimer’s disease (AD), they are quite often underrecognized. ¹ In the early stages of AD, various nutritional problems may occur leading to reduced dietary intake, weight loss, and undernutrition. ^{2 –4} The prevalence of weight loss in older patients with AD is high, ⁵ estimated to affect approximately 25% to 45% of patients ^3,6 and increases directly with disease severity. ⁷ Involuntary weight loss is a highly frequent preclinical marker in AD, ^6,8,9 has a strong negative impact on health status and quality of life, ^3,6,10 and has been associated with rapid functional ^11,12 and cognitive decline ^12,13 in community-dwelling patients with AD. ^{10 –14}

Some studies have shown that circa 3% of mild to moderate ⁸ and half of patients with severe AD were undernourished. ¹ Moreover, a recent review by Vandewoude et al ³ showed that 26% to 80% of community-dwelling patients with AD were at risk of undernutrition by the Mini Nutritional Assessment (MNA) score. This high prevalence is worrying because undernourished patients with AD are more prone to adverse outcomes, such as increased risk of loss of muscle mass, ¹² reduced muscle strength and muscle function, ¹² impaired autonomy, ⁸ worse cognition, ^8,12 higher rates of infection, ¹² skin ulcers, ¹² increased burden of disease, ^12,15,16 and use of health-care resources. ¹⁶ Taking this into account, the screening, diagnosis, and treatment of undernutrition in community-dwelling patients with AD must be a priority for health professionals.

A systematic review from 2013 by Allen et al ¹⁵ showed that providing oral nutritional supplementation (ONS) has a positive effect on weight gain and cognition at follow-up in older people with dementia. Recent guidelines on nutrition in dementia by the European Society for Clinical Nutrition and Metabolism ² recommended ONS to improve nutritional status but not to ameliorate cognitive impairment or prevent cognitive decline. Indeed, a considerable body of evidence resulting from experimental studies has shown that there is a short-term positive effect of ONS on nutritional status as assessed by the MNA score ^17,18 or through anthropometrical indicators such as body weight, ^{17 –20} body mass index (BMI), ^18,19,21 triceps skinfold thickness (TST), ^18,19,21 mid-upper arm circumference, ^19,21 mid-upper arm muscle circumference (MUAMC), ¹⁹ and fat-free mass (FFM). ¹⁷ Additionally, randomized controlled trials showed that a physical exercise program ²² or a teaching and training intervention ²³ for patients with AD ²² or dementia ²³ had a positive effect on their nutritional status, MNA score, ²³ BMI, ²³ and body weight, ²³ and also on the Katz index of activities of daily living (ADLs). ²²

It was consistently reported by Manckoundia et al, ²⁴ Mazoteras Munoz et al, ²⁵ and Gras et al ²⁶ that 9% to 52% of patients with AD have gait and balance disorders. ²⁴ Gait speed (GS) impairment is a clinical marker of AD, ²⁷ and an early physical therapy intervention may slow gait disorders improving functional independence. ²⁶ While these results are promising, it is of major relevance to explore whether these strategies are effective to prevent nutritional and functional status impairment or cognitive decline.

However, the effectiveness of an ONS intervention associated with a psychomotor rehabilitation program on nutritional, functional, and cognitive parameters in undernourished or nutritionally at-risk community-dwelling patients with AD remains to be seen. Therefore, the present study aims to investigate the impact of a 21-day ONS and a psychomotor rehabilitation program on nutritional status and body composition, physical performance (muscle function), muscle strength, and cognitive status in probable mild community-dwelling patients with AD at 6-month follow-up.

Participants and Methods

Sample and Study Design

A prospective randomized controlled study was conducted in outpatients with AD from a psychogeriatric department of a psychiatric hospital in an intervention group and in a control group (CG). A third intervention group from an AD day care center was formed. A 21-day intervention program and a follow-up, which lasted for 6 months from baseline, were conducted. Outpatients with AD from the psychogeriatric department were consecutively recruited and randomized into 2 groups: the nutritional supplementation group (NSG; n = 25) to receive once a day a small volume energy-dense liquid ONS (125 mL, 2.4 kcal/mL) plus standard dietetic advice (SDA) and the CG (n = 43) that received only SDA. A third intervention group was constituted of patients from an AD day care center, where participants were recruited on a convenience basis. This nutritional supplementation psychomotor rehabilitation group (NSPRG; n = 11) received the same 21-day ONS, SDA, and also a customized psychomotor rehabilitation program.

Eligibility criteria were set as the existence of a diagnosis of probable AD according to the National Institute of Neurological and Communicative Diseases and Stroke/Alzheimer Disease and Related Disorders Association criteria, ²⁸ and dementia was defined according to the Diagnostic Statistics Mental Disorders (Fifth Edition). ²⁹ To confirm the diagnosis of dementia, a detailed history and examination by the primary care physician was required, including a comprehensive assessment of the patient’s symptoms and concerns and their social functioning. Neurological and psychiatric examinations, including standard laboratory tests and magnetic resonance tomography or computed tomography, were also performed. A trained psychiatrist then confirmed the diagnosis of AD for all participants. Participants were considered to have mild to moderate dementia if Mini-Mental State Examination (MMSE) score was ≥10 points and ≤26 points.

Moreover, only participants with 65 years of age or more, presenting weight loss higher than 5% of body weight in the previous year, being at risk of undernutrition or undernourished according to MNA score classification, having an active caregiver, living at home with their family, and willing to commit to the 6-month study period were recruited.

Exclusion criteria were blindness and/or deafness, no identified caregiver, severe acute illness or a known terminal illness, diagnosis of cancer in the last 5 years, enteral or parenteral nutritional support, dietary advice or use of ONS in the preceding month, participation in a rehabilitation program, unable to walk for at least 4 m without assistance, use of a gait aid and being unable to follow instructions concerning the testing procedure for GS, and the handgrip strength (HGS) measurements.

All 87 participants who fulfilled the inclusion criteria were invited to participate. Of these, 8 (9.1%) were not included because they did not complete the trial, 1 of the 26 patients enrolled in the NSG died, 3 of the 15 patients enrolled in the NSPRG dropped out of the study, and 1 was admitted to a nursing home; of the 46 patients enrolled in the CG, 1 dropped out at the beginning of the study and 2 died before a third evaluation.

This research was conducted according to the Declaration of Helsinki ²⁹ and was approved by the institutional ethics review board (number 9/2011). All study participants and their legal tutors gave their informed consent.

Nutritional Intervention

During the study, participants from the NSG and the NSPRG consumed daily at midmorning a small volume high-protein energy-dense liquid ONS (125 mL) containing 300 kcal/d (12 g protein, 37.1 g carbohydrates, and 11.6 g fat). This ONS was available in 4 flavors (strawberry, chocolate, vanilla, and coffee) and was given to each participant in the original packaging, with a straw. The caregivers kept an adherence diary where they recorded the intake amount of ONS immediately after consumption.

Psychomotor Rehabilitation Program

For the psychomotor rehabilitation program, participants of the NSPRG were divided into 2 different session schedules according to their baseline psychomotor performance scores, MMSE score, behavior disturbances, and affinity between participants. Although in the European countries this program is commonly called as neuropsychomotricity and used to stimulate children affected by neurodevelopmental disorders, ^30,31 a specific geriatric psychomotor rehabilitation program was developed for patients with AD, including a multicomponent/modular therapy program with targeted objectives, consisting of attentional tasks, strength, tonicity, static and dynamic balance exercises, body awareness, spatial and temporal restructuration, immediate and working memories and praxis, fine motor skills, and gross motor skills. ^32,33 The sessions consisted of 1 hour × twice week exercises in the morning and each participant completed 12 sessions. A psychomotor therapist conducted all the sessions.

All the participants from the studied groups consumed their regular diet and received SDA from the same nutritionist. The aim was to promote adequate energy and nutrient intake by addressing issues such as information about the importance of the nutrition in AD, behavioral changes during meals, loss of appetite, and how to detect swallowing difficulties. After discontinuing intervention on the 21st day, all participants were followed up and were reevaluated at 90 and 180 days since baseline (Figure 1).

Figure 1.

Study design.

Data Collection

Nutritional, functional, and cognitive status parameters were considered in order to compare the groups at baseline and to measure outcomes. Nutritional status was evaluated through body weight, BMI, TST, MUAMC, FFM, FFM index, fat mass, phase angle (PA), and the MNA full-form score.

Body weight (kg) was determined with a portable calibrated digital scale, with a resolution of 0.1 kg (Seca, model 7701321004, Vogel & Hamburg, Germany, model 770 1321004, Germany), with patients wearing light clothes and barefoot. Due to the difficulty in obtaining reliable height measurements, height was obtained from the value recorded in their national identity card. The TST (mm) was measured with a Lange caliper^® (Beta Technology Incorporated, Cambridge, Maryland) with a resolution of 0.2 mm. ³⁴ Muscle mass was estimated through MUAMC (cm) and assessed by bioelectrical impedance analysis (Physiological Data Analyzer System; Akern 101, Florence, Italy). All anthropometric measurements were performed by the same trained researcher using standard methods. ³⁵ Bioelectrical impedance analysis was assessed on the right side of the body in the supine position. ³⁶ The PA (degrees) and FFM were calculated. ³⁷ According to the MNA total score, patients were categorized as undernourished (MNA < 17) or at risk of undernutrition (MNA = 17.1-23.5). ³⁸

Muscle strength was measured using a calibrated TTM, ORIGINAL SMEDLAY’S DYNAMO METER (100kg Tokyo, resolution of 0.5 Kgf), according to the American Society of Hand Therapists protocol. ³⁹ The HGS measurements were taken with the participant in the sitting position and with the arm by the side of the body and the forearm stretched to an angle of 90°. Each participant performed 1 test trial and then 3 test measurements were taken with a 1-minute interval between them, and the maximum value was recorded. Muscle function was evaluated by GS (m/s) at usual pace, using the cutoffs proposed by the International Academy on Nutrition and Aging expert panel ⁴⁰ : GS ≤ 0.8 m/s (5 seconds to perform a 4-m course) for both men and women.

Autonomy in ADLs was assessed with the Barthel index (BI). ⁴¹ The maximum score is 100 and the minimum is 0 (total dependence). Independence in basic and instrumental ADLs (IADLs) was evaluated with the Lawton and Brody scale. ⁴² The higher the score, the worse the autonomy and the independence. A score above 20 indicates a severe dependence and the minimum score is 8, signifying independence.

Cognitive impairment was evaluated through the clock drawing test (CDT) ^43,44 and the MMSE tool score. ⁴⁵ For the CDT, participants were asked to draw a clock face and place all the numbers on it, and upon completion “make the clock say 10 minutes after 11.” The clock face was scored by dividing it into 8s, beginning with a line through the number 12 and the center of the circle. One point was given for placing each of the numbers 1, 2, 4, 5, 7, 8, 10, and 11 in the proper octant of the circle, and 1 point each was given for drawing a short hand pointing to the 11 and a long hand pointing to the 2. A normal score ranges between 7 and 9 points. A score of 0 to 3 points was considered to be indicative of clinically significant cognitive impairment. The MMSE screening assesses orientation, memory, and other cognitive skills, providing a total score ranging from 0 to 30, with higher scores indicating better cognitive status. A score of <20 points is usually considered to be indicative of clinically significant cognitive impairment. The previously validated normative cutoff values for the Portuguese population adjusted to education level were used ⁴⁶ : zero years of schooling (MMSE ≤ 15), 1 to 11 years of schooling (MMSE ≤ 22), and more than 11 years of schooling (MMSE ≤ 27).

Data Analysis

The Kolmogorov-Smirnov test was used to evaluate the normality of the variables distribution. Means (standard deviations [SD]) were calculated for continuous variables. Categorical variables were reported as frequencies. Differences in categorical data were evaluated by Pearson χ² test or Fisher exact test. Comparison of baseline characteristics between the 3 groups and the differences between groups after intervention were carried out using the Mann-Whitney U test, the independent samples Student t-test, or the Wilcoxon test, as appropriate. Changes in MNA score (▵MNA) between NSG, NSPRG, and CG at 21, 90, and 180 days were calculated as follows: 21 days minus baseline, 90 days minus baseline, and 180 days minus baseline. All ▵MNA at 21, 90, and 180 days values were log-transformed (base-e logarithm), prior to mixed and general linear model (GLM) analysis. The GLM method was used to examine the differences in ▵MNA at 21, 90, and 180 days for NSG, NSPRG, and CG, respectively. In this analysis, NSG and NSPRG were coded as dummy variables and the control condition was used as the reference group. Main effects were examined among NSG and NSPRG.

To examine potential ▵MNA differences in intervention groups (NSG and NSPRG), the MNA and MMSE score effects were explored at baseline (fixed factors) and by adding 2 interaction terms (NSG × MNA at baseline and NSG × MMSE at baseline vs NSPRG × MNA at baseline and NSPRG × MMSE at baseline). All P values were adjusted using the Dunnett method for multiple pairwise comparisons ⁴⁷ or the Bonferroni method as appropriate. The α levels were set at .05. Statistical analyses were carried out with the Software Package for Social Sciences (SPSS) for Windows (version 24.0; SPSS Inc, an IBM Company, Chicago, Illinois).

Results

This study included 79 participants with AD, with a mean age of 78.2 (SD = 6.6) years (ranging from 65 to 93 years). According to the MNA score, 55.7% of the study participants were undernourished, while the remaining sample was at nutritional risk. The mean CDT score showed significant cognitive impairment (96.2%) and the mean MMSE revealed moderate dementia (55.7%). The mean BI and IADL Lawton scale scores presented moderate autonomy (78.5%) and severe dependence in ADLs (77.2%), respectively. Descriptive statistics of NSG, NSPRG, and CG are presented in Table 1. These groups are homogeneous regarding the studied characteristics, except for education with NSG participants having less schooling years than NSPRG (P = .019) and NSPRG revealing a higher MNA score (P = .026). In fact, NSG and CG presented worse nutritional status, with more than half of the participants being undernourished (respectively, 52% and 65.1%), whereas a lower proportion (27.3%) of the NSPRG patients were undernourished (Table 1). The compliance with the ONS and to the psychomotor rehabilitation program was excellent, without any refusals or dropouts in both intervention groups, the NSG and the NSPRG.

Table 1.

Baseline Characteristics of the Nutritional Supplementation Group (NSG), Nutritional Supplementation Psychomotor Rehabilitation Group (NSPRG), and Control Group (CG).^a

Variables	NSG (n = 25)	NSPRG (n = 11)	CG (n = 43)	P b	P c	P d	P e
Age, years, mean (SD)	77.8 (7.2)	80.0 (6.4)	78.0 (6.4)	.627	.985	.608	.506
Sex (males/females)	10/15	7/4	15/28	.229	.893	.162	.281
Education (school years), mean (SD)	3.9 (2.1)	6.9 (4.0)	5.4 (3.4)	.027	.101	.335	.019
Clock drawing test score, mean (SD)f	1.1 (1.7)	2.1 (1.8)	1.2 (1.4)	.296	.983	.267	.220
Mini-Mental State Examination score, mean (SD)f	18.6 (4.9)	19.3 (5.4)	20.0 (4.9)	.557	.476	.899	.879
Barthel index score, mean (SD)f	76.4 (19.7)	86.3 (14.5)	78.7 (18.3)	.325	.847	.385	.211
Independence, n (%)	4 (16.0)	3 (27.3)	7 (16.3)
Moderate dependence, n (%)	20 (80.0)	8 (72.7)	34 (79.1)
Severe dependence, n (%)	1 (4.0)	0.0 (0.0)	2 (4.7)
Lawton index score, mean (SD)g	24.8 (5.2)	28.3 (2.5)	24.3 (5.6)	.081	.939	.052	.099
Independence, n (%)	0.0 (0.0)	0.0 (0.0)	1 (2.3)
Moderate dependence, n (%)	6 (24.0)	0.0 (0.0)	11 (25.6)
Severe dependence, n (%)	19 (76.0)	11 (100)	31 (72.1)
Weight, kg, mean (SD)	63.4 (15.4)	70.3 (13.3)	63.3 (14.7)	.353	.999	.293	.300
Mini Nutritional Assessment score, mean (SD)h	15.2 (4.3)	18.5 (2.5)	15.3 (3.4)	.031	.990	.024	.026
At risk of undernourished, n (%)	12 (48.0)	8 (72.7)	15 (34.9)
Undernourished, n (%)	13 (52.0)	3 (27.3)	28 (65.1)
Triceps skinfold thickness, mm, mean (SD)	16.1 (10.4)	14.0 (6.0)	17.3 (10.6)	.630	.875	.565	.754
Fat-free mass index, kg/m2, mean (SD)	16.2 (2.4)	17.5 (2.1)	16.8 (1.9)	.271	.569	.510	.173
Women, mean (SD)	15.2 (1.6)	15.2 (1.4)	15.7 (1.5)	.590
Men, mean (SD)	17.8 (2.7)	18.8 (1.1)	18.7 (0.9)	.164
Handgrip strength, kgf, mean (SD)	15.9 (9.1)	16.6 (6.7)	14.1 (7.1)	.516	.588	.570	.946
Women	9.8 (4.5)	11.3 (3.7)	11.0 (5.3)
Men	25.1 (6.0)	19.7 (6.1)	20.1 (6.4)
Gait speed, m/s, mean (SD)	9.5 (4.2)	9.0 (2.8)	10.7 (4.8)	.392	.489	.443	.908

Abbreviation: SD, standard deviation.

^aP values were adjusted using Dunnett method.

^bSignificant differences between NSG, NSPRG, and CG.

^cSignificant differences between NSG and CG.

^dSignificant differences between NSPRG and CG.

^eSignificant differences between NSPRG and NSG.

^fHigher scores represent better function.

^gHigher scores represent worse function.

^hHigher scores represent better nutritional status.

Differences between the NSG, NSPRG, and CG at 21, 90-, and 180 days for nutritional parameters are displayed in Table 2. These differences for functional and cognitive status are presented in Table 3 and described below. Table 4 shows differences in ▵MNA score at 21, 90, and 180 days. Mixed and GLM testing of the effects of the MNA at baseline, MMSE at baseline, NSG × MNA at baseline, NSG × MMSE at baseline, NSPRG × MNA at baseline, and NSPRG × MMSE at baseline on ▵MNA at 21, 90, and 180 days are presented in Table 5. Primary end point was ▵MNA from baseline at 21 days. Secondary end points were ▵MNA from baseline at 90 and 180 days. Significant mean differences for the ▵MNA between NSPRG and NSG were observed at 21 days (P = .002) and at 180 days (P = .007; Table 4).

Table 2.

Nutritional Parameters Differences Between Nutritional Supplementation Group (NSG), Nutritional Supplementation Psychomotor Rehabilitation Group (NSPRG), and Control Group (CG) at Baseline (T0), 21 Days (T21D), 90 Days (T90D), and 180 Days (T180D).^a,b

Variables	T0	T21D	P c	P d	P e	P f	T90D	P g	P h	P i	P j	T180D	P k	P l	P m	P n
Weight, kg
NSG	63.4 (15.4)	63.7 (15.0)					64.1 (15.7)					64.3 (15.6)
NSPRG	70.3 (13.3)	71.9 (14.2)					71.8 (14.7)					70.6 (13.4)
CG	63.3 (14.7)	63.4 (13.9)					63.3 (14.3)					64.1 (13.5)
Weight changes, kgo
NSG		0.3 (1.4)		.338	.895		0.6 (2.5)		.183	.541		0.8 (3.2)		.276	1
NSPRG		1.5 (1.2)		.005		.025	1.5 (2.4)		.041		.183	0.2 (2.4)		.722		.874
CG		0.1 (1.8)		.643			0.0 (2.6)		.562			0.8 (3.7)		.122
BMI, kg/m2
NSG	24.1 (5.3)	24.3 (5.3)					24.4 (5.4)					24.5 (5.3)
NSPRG	24.9 (4.2)	25.5 (4.6)					25.5 (4.7)					25.0 (4.1)
CG	24.3 (5.6)	24.4 (5.3)					24.3 (5.5)					24.7 (5.2)
BMI changes, kg/m2o
NSG		0.1 (0.5)		.346	.882		0.2 (0.9)		.158	.527		0.3 (1.2)		.397	.989
NSPRG		0.5 (0.4)		.005		.053	0.5 (0.8)		.041		.256	0.0 (0.8)		.182		.809
CG		0.0 (0.7)		.629			0.0 (1.0)		.681			0.3 (1.5)		.132
MNA (score)
NSG	15.2 (4.3)	15.6 (4.6)					16.5 (3.9)					17.0 (3.6)
NSPRG	18.5 (2.5)	20.0 (2.4)					20.1 (2.3)					18.2 (2.7)
CG	15.3 (3.4)	15.1 (3.3)					15.6 (3.3)					17.2 (3.7)
MNA (score) changeso
NSG		0.4 (0.8)		.011	.045		1.3 (1.2)		<.001	.029		1.8 (2.0)		<.001	1
NSPRG		1.5 (1.0)		.007		<.001	1.6 (1.0)		.007		.028	−0.2 (1.9)		.766		.018
CG		−0.1 (1.1)		.317			0.3 (1.7)		.094			1.8 (2.5)		<.001
MNA (score), n (%)			<.001					.004					.872
NSG
Nonundernourishedc	0	1 (4.0)					0					1 (4.0)
At risk of undernourishedc	12 (48.0)	13 (52.0)					13 (52.0)					12 (48.0)
Undernourishedc	13 (52.0)	11 (44.0)					12 (48.0)					12 (48.0)
NSPRG
Nonundernourishedc	0	0					0					0
At risk of undernourishedc	8 (72.7)	10 (90.9)					10 (90.9)					7 (63.6)
Undernourishedc	3 (27.3)	1 (9.1)					1 (9.1)					4 (36.4)
CG
Nonundernourishedc	0	0					0					1 (2.3)
At risk of undernourishedc	15 (34.9)	14 (32.6)					17 (39.5)					25 (58.1)
Undernourishedc	28 (65.1)	29 (67.4)					26 (60.5)					17 (39.5)
TST (mm)
NSG	16.1 (10.4)	16.1 (9.8)					16.4 (9.2)					18.0 (10.6)
NSPRG	14.0 (6.0)	16.5 (8.7)					16.2 (7.7)					16.9 (8.7)
CG	17.3 (10.6)	18.0 (9.1)					18.2 (9.4)					18.7 (9.9)
TST changes, mmo
NSG		−0.0 (4.4)		.630	.770		0.3 (5.3)		.636	.862		1.8 (6.5)		.075	.968
NSPRG		2.4 (3.9)		.027		.509	2.1 (3.2)		.057		.710	2.8 (3.5)		.015		.769
CG		0.7 (5.3)		.505			0.9 (5.2)		.371			1.5 (5.9)		.051
MUAMC, cm
NSG	20.5 (2.0)	20.4 (1.9)					20.0 (1.5)					19.7 (2.0)
NSPRG	21.9 (2.2)	20.9 (1.8)					21.0 (2.2)					20.4 (2.0)
CG	20.6 (3.1)	20.2 (2.9)					20.3 (2.9)					20.0 (2.5)
MUAMC changes, cmo
NSG		−0.0 (1.3)		.872	.732		−0.4 (1.5)		.424	.702		−0.8 (2.4)		.050	.843
NSPRG		−0.9 (1.0)		.018		.414	−0.9 (0.9)		.024		.286	−1.4 (1.0)		.004		.374
CG		−0.3 (1.9)		.280			−0.2 (1.5)		.431			−0.5 (2.3)		.120
FFM, kg
NSG	43.0 (9.0)	42.6 (7.8)					42.8 (8.1)					43.5 (8.4)
NSPRG	49.9 (11.0)	50.3 (12.1)					49.9 (10.8)					49.4 (10.9)
CG	43.9 (7.4))	43.7 (6.9)					44.4 (6.7)					44.2 (7.1)
FFM changes, kgo
NSG		−0.3 (1.9)		.619	.862		−0.1 (2.1)		.527	.451		0.2 (2.2)		.657	.972
NSPRG		0.4 (2.2)		.722		.738	−0.0 (2.8)		.859		.773	−0.4 (2.1)		.534		.499
CG		−0.1 (2.4)		.131			0.5 (2.8)		.035			0.3 (2.6)		.288
FFM, %
NSG	69.2 (12.0)	68.5 (12.5)					68.6 (13.2)					68.9 (12.4)
NSPRG	71.3 (10.0)	70.2 (10.4)					69.8 (8.0)					70.4 (10.3)
CG	69.8 (11.7)	70.3 (11.9)					71.5 (12.5)					70.2 (11.8)
FFM changes, %o
NSG		−0.7 (2.0)		.069	.342		−0.6 (3.1)		.326	.070		−0.3 (3.1)		.882	.777
NSPRG		−1.0 (2.3)		.155		.385	−1.5 (3.2)		.075		.063	−0.9 (2.9)		.424		.649
CG		0.4 (4.5)		.388			1.7 (5.1)		.082			0.3 (5.7)		.391
FFMI, kg/m2
NSG	16.2 (2.4)	16.1 (2.0)					16.2 (2.0)					16.4 (2.2)
NSPRG	17.5 (2.1)	17.6 (2.4)					17.5 (2.2)					17.3 (2.0)
CG	16.8 (1.9)	16.7 (1.8)					17.0 (1.8)					16.9 (1.7)
FFMI changes, kg/m2o
NSG		−0.1 (0.6)		.638	.838		−0.0 (0.8)		.459	.391		0.1 (0.8)		.600	.985
NSPRG		0.1 (0.7)		.722		.825	0.0 (1.1)		.859		.764	−0.1 (0.7)		.594		.521
CG		−0.0 (0.8)		.131			0.2 (1.0)		.032			0.1 (1.0)		.288
FM, %
NSG	30.7 (12.0)	31.4 (12.5)					31.3 (13.2)					31.0 (12.4)
NSPRG	28.6 (10.0)	29.7 (10.4)					30.1 (8.0)					29.5 (10.3)
CG	30.1 (11.7)	29.9 (11.5)					28.4 (12.5)					29.7 (11.8)
FM changes, %o
NSG		0.7 (2.0)		.069	.482		0.6 (3.1)		.326	.070		0.3 (3.1)		.882	.777
NSPRG		1.0 (2.3)		.155		.491	1.5 (3.2)		.075		.063	0.9 (2.9)		.424		.649
CG		−0.2 (4.4)		.718			−1.7 (5.1)		.082			−0.3 (5.7)		.391
PA, degrees
NSG	3.4 (0.5)	3.5 (0.6)					3.4 (0.6)					3.3 (0.7)
NSPRG	3.5 (0.5)	3.5 (0.5)					3.7 (0.7)					3.2 (0.5)
CG	3.5 (0.8)	3.5 (0.8)					3.4 (0.7)					3.4 (0.8)
PA changes, degreeso
NSG		0.0 (0.2)		.543	.345		−0.0 (0.4)		.367	.910		−0.0 (0.4)		.367	.970
NSPRG		0.0 (0.2)		1.000		.535	0.2 (0.6)		.575		.206	−0.2 (0.6)		.248		.786
CG		−0.0 (0.2)		.466			−0.0 (0.5)		.124			−0.1 (0.5)		.044

Abbreviations: BMI, body mass index; FM, fat mass; FFM, fat-free mass; FFMI, fat-free mass index; MNA, Mini Nutritional Assessment; MUAMC, mid-upper arm muscle circumference; PA, phase angle; TST, triceps skinfold thickness.

^aAll data are expressed as means (standard deviations), except for those with,^c which are expressed as number and percentage.

^bP values were adjusted using Dunnett method.

^cSignificance differences between NSG, NSPRG, and CG.

^dSignificance differences in each group (NSG, NSPRG, and CG) between T21D and T0 (Wilcoxon test).

^eSignificance differences between NSG and CG at T21D minus T0 (Mann-Whitney U test).

^fSignificance differences between NSPRG and CG at T21D minus T0 (Mann-Whitney U test).

^gSignificance differences between NSG, NSPRG, and CG.

^hSignificance differences in each group (NSG, NSPRG, and CG) between T90D and T0 (Wilcoxon test).

ⁱSignificance differences between NSG and CG at T90D minus T0 (Mann-Whitney U test).

^jSignificance differences between NSPRG and CG at T90D minus T0 (Mann-Whitney U test).

^kSignificance differences between NSG, NSPRG, and CG.

^lSignificance differences in each group (NSG, NSPRG, and CG) between T180D and T0 (Wilcoxon test).

^mSignificance differences between NSG and CG at T180D minus T0 (Mann-Whitney U test).

ⁿSignificance differences between NSPRG and CG at T180D minus T0 (Mann-Whitney U test).

^oMean (standard deviations) of the differences were calculated as: T21D minus T0, T90D minus T0, and T180D minus T0.

Table 3.

Functional and Cognitive Parameters Differences Between Nutritional Supplementation Group (NSG), Nutritional Supplementation Psychomotor Rehabilitation Group (NSPRG), and Control Group (CG) at Baseline (T0), 21 Days (T21D), 90 Days (T90D), and 180 Days (T180D).^a,b

Variables	T0	T21D	P c	P d	P e	T90D	P f	P g	P h	T180D	P i	P j	P k
HGS, kgf
NSG	15.9 (9.1)	15.4 (9.2)				15.4 (8.9)				14.4 (8.9)
NSPRG	16.6 (6.7)	17.5 (7.1)				17.8 (6.9)				16.7 (6.3)
CG	14.1 (7.1)	13.2 (6.7)				13.4 (7.0)				14.1 (9.7)
HGS changes, kgfl
NSG		−0.4 (2.3)	.376	.810		−0.5 (2.9)	.558	.959		−1.4 (3.3)	.080	.527
NSPRG		0.9 (5.4)	.477		.161	1.1 (5.5)	.722		.204	0.1 (5.8)	.756		.994
CG		−0.9 (2.8)	.062			−0.7 (3.2)	.164			−0.0 (6.4)	.254
Gait speed, m/s
NSG	9.5 (4.2)	9.6 (3.7)				9.3 (3.8)				9.0 (3.5)
NSPRG	9.0 (2.8)	7.3 (2.0)				7.2 (2.1)				9.7 (3.6)
CG	10.7 (4.8)	10.8 (4.6)				11.5 (5.1)				11.1 (4.3)
Gait speed changes, m/sl
NSG		0.1 (2.8)	.275	1		−0.2 (2.9)	.439	.279		−0.4 (3.6)	.687	.566
NSPRG		−1.6 (1.8)	.013		.056	−1.7 (2.6)	.063		.022	0.7 (4.4)	.539		.976
CG		0.1 (2.0)	.512			0.8 (3.0)	.055			0.4 (3.8)	.107
Barthel index (score)
NSG	76.4 (19.7)	76.4 (19.7)				76.4 (19.8)				76.1 (19.8)
NSPRG	86.3 (14.5)	86.3 (14.5)				86.3 (14.5)				86.3 (14.5)
CG	78.7 (18.3)	78.3 (18.8)				78.4 (18.8)				78.0 (18.4)
Barthel index (score) changesl
NSG		0.0 (0.0)	1.000	.704		0.0 (1.4)	1.000	.898		−0.2 (2.2)	.581	.738
NSPRG		0.0 (0.0)	1.000		.822	0.0 (0.0)	1.000		.942	0.0 (0.0)	1.000		.707
CG		−0.3 (2.5)	.414			−0.2 (2.8)	.593			−0.6 (3.3)	.202
Lawton index (score)
NSG	24.8 (5.2)	24.8 (5.2)				25.1 (5.2)				25.1 (5.2)
NSPRG	28.3 (2.5)	28.3 (2.5)				28.3 (2.5)				28.3 (2.5)
CG	24.3 (5.6)	24.6 (5.6)				24.6 (5.7)				24.7 (5.5)
Lawton index (score) changesl
NSG		0.0 (0.0)	1.000	.499		0.3 (1.6)	.317	.992		0.3 (1.6)	.655	.987
NSPRG		0.0 (0.0)	1.000		.636	0.0 (0.0)	1.000		.805	0.0 (0.0)	1.000		.703
CG		0.2 (1.2)	.180			0.2 (1.5)	.285			0.3 (1.6)	.684
MMSE (score)
NSG	18.6 (4.9)	18.4 (5.2)				18.4 (5.2)				17.5 (6.6)
NSPRG	19.3 (5.4)	18.0 (6.6)				18.0 (6.6)				17.6 (8.4)
CG	20.0 (4.9)	20.0 (4.9)				20.0 (4.9)				20.0 (4.9)
MMSE (score) changesl
NSG		−0.2 (1.3)	1.000	.654		−0.2 (1.3)	1.000	.700		−1.1 (3.2)	1.000	.162
NSPRG		−1.2 (3.1)	1.000		.015	−1.2 (3.1)	1.000		.017	−1.7 (5.1)	.317		.108
CG		0.0 (0.0)	1.000			−0.0 (0.1)	1.000			−0.0 (0.1)	.317

Abbreviations: HGS, handgrip strength; MMSE, Mini-Mental State Examination.

^aAll data are expressed as means (standard deviations).

^bP Values were adjusted using Dunnett method.

^cSignificance differences in each group (NSG, NSPRG, and CG) between T21D and T0 (Wilcoxon test).

^dSignificance differences between NSG and CG at T21D minus T0 (Mann-Whitney U test).

^eSignificance differences between NSPRG and CG at T21D minus T0 (Mann-Whitney U test).

^fSignificance differences in each group (NSG, NSPRG, and CG) between T90D and T0 (Wilcoxon test).

^gSignificance differences between NSG and CG at T90D minus T0 (Mann-Whitney U test).

^hSignificance differences between NSPRG and CG at T90D minus T0 (Mann-Whitney U test).

ⁱSignificance differences in each group (NSG, NSPRG, and CG) between T180D and T0 (Wilcoxon test).

^jSignificance differences between NSG and CG at T180D minus T0 (Mann-Whitney U test).

^kSignificance differences between NSPRG and CG at T180D minus T0 (Mann-Whitney U test).

^lMean (standard deviations) of the differences were calculated as: T21D minus T0, T90D minus T0, and T180D minus T0.

Table 4.

Differences in the Change Mini Nutritional Assessment Score (ΔMNA) Between Nutritional Supplementation Group (NSG) and Nutritional Supplementation Psychomotor Rehabilitation Group (NSPRG) at 21, 90, and 180 Days.

	NSG (n = 25)	NSPRG (n = 11)	Mean Difference	95% CI	p a
▵MNA at 21 days, mean (SD)b	0.4 (0.8)	1.5 (1.0)	1.15	0.4 to 1.8	.002
▵MNA at 90 days, mean (SD)b	1.3 (1.2)	1.6 (1.0)	0.33	−0.5 to 1.1	.432
▵MNA at 180 days, mean (SD)b	1.8 (2.0)	−0.2 (1.9)	−2.10	−3.6 to −0.6	.007

Abbreviations: CI, CI, confidence interval; SD, standard deviation.

^aSignificance differences in the ΔMNA between NSPRG and NSG.

^bΔMNA was calculated as: 21-day minus baseline, 90-day minus baseline, and 180-day minus baseline. P Value (Mann-Whitney U test).

Table 5.

Results From the Generalized Linear Models Testing the Effects of the MNA at Baseline (B), MMSE at Baseline (C), (A × B), (A × C), (D × B), (D × C), on Changes MNA (ΔMNA) at 21 Days (T21D), 90 Days (T90D), and 180 Days (T180D) of NSG (A) and NSPRG (D).^a,b

Source of Variation	χ2	df	Mean Square	F	p	χ2	df	Mean Square	F	p	χ2	df	Mean Square	F	p
NSG (A)	0.20	1	0.20	0.46	.510	0.00	1	0.00	0.01	.890	0.39	1	0.39	0.62	.433
Intercept	0.01	1	0.01	0.03	.864	0.92	1	0.92	1.90	.175	4.41	1	4.41	7.05	.011
MNA at baseline (B)	0.44	1	0.44	0.98	.342	0.24	1	0.24	0.49	.486	4.27	1	4.27	6.83	.012
MMSE at baseline (C)	0.65	1	0.65	1.44	.254	0.11	1	0.11	0.24	.627	0.26	1	0.26	0.42	.517
A × B	0.25	1	0.25	0.56	.466	0.00	1	0.00	0.00	.936	0.00	1	0.00	0.01	.920
A × C	0.02	1	0.02	0.05	.827	0.01	1	0.01	0.03	.853	0.67	1	0.67	1.07	.305
Error	4.93	11	0.44			18.87	39	0.48			31.27	50	0.62
Total	9.31	17				26.40	45				63.12	56
NSPRG (D)	0.20	1	0.20	0.46	.510	0.48	1	0.48	1.11	.297	0.32	1	0.32	0.51	.475
Intercept	0.01	1	0.01	0.03	.864	0.00	1	0.00	0.00	.987	0.66	1	0.66	1.03	.314
MNA at baseline (B)	0.44	1	0.44	0.98	.342	0.11	1	0.11	0.26	.608	0.54	1	0.54	0.86	.357
MMSE at baseline (C)	0.65	1	0.65	1.44	.254	0.03	1	0.03	0.07	.784	0.02	1	0.02	0.04	.835
D × B	0.25	1	0.25	0.56	.466	0.78	1	0.78	1.79	.188	0.22	1	0.22	0.35	.553
D × C	0.02	1	0.02	0.05	.827	0.00	1	0.00	0.01	.907	0.14	1	0.14	0.23	.632
Error	4.93	11	0.44			16.96	39	0.43			31.81	50	0.63
Total	9.31	17				26.40	45				63.12	56

Abbreviations: χ², chi-square test; df, degrees of freedom; F, 1-way univariate analysis of variance (ANOVA, the F test); MNA, Mini Nutritional Assessment; MMSE, Mini-Mental State Examination; NSG, nutritional supplementation group; NSPRG, nutritional supplementation psychomotor rehabilitation group.

^aP < .05.

^bP values were adjusted by the Bonferroni method.

^cNSG, R² = .346, adjusted R² = −.048.

^dNSPRG, R² = .173, adjusted R² = .054.

^eNSG, R² = .033, adjusted R² = −.090.

^fNSPRG, R² = .131, adjusted R² = .020.

^gNSG, R² = .184, adjusted R² = .102.

^hNSPRG, R² = .169, adjusted R² = .086.

Nutritional, Functional, and Cognitive Changes From Baseline to 21 Days

The 21-day ONS resulted in a significant improvement in the NSG and NSPRG participants’ nutritional status (Table 2). Compared to CG, the MNA score increased (mean [SD]) 0.4 [0.8] points in the NSG and 1.5 [1.0] points in the NSPRG, P < .05; Table 2). Seventy-six percent of the NSG participants improved their MNA score (from 0.5 to 3 points), while in the NSPRG, 81.9% of the participants increased the MNA score from 1 to 3.5 points, decreasing the frequency of undernourished patients from 27.3% to 9.1%. Furthermore, improvements in body weight of 1.5 (1.2) kg (P < .05) were also observed in the NSPRG compared with the CG. Nearly all the NSPRG participants (81.8%) increased their body weight, with this increment ranging between 1.1 and 3.8 kg. Nonetheless, in the CG, an increase in the proportion of undernourished patients from 65.1 to 67.4 was found.

No significant changes were found for functional status in this period. Concerning the cognitive status, a slight decrease in the MMSE score was also observed in the NSG and in the NSPRG of −0.2 (1.3) and −1.2 (3.1) points, respectively (P < .05; Table 3).

Nutritional, Functional, and Cognitive Changes From Baseline to 90 Days and to 180 Days

At 90-day follow-up, no significant changes for the functional status were noticed in the NSG, compared to the CG (Table 3). Among NSPRG, 81.8% of the participants increased their baseline GS. This increase ranged from −1 to −4 m/s to complete the 4-m test, with a mean value of −1.7 (2.6) m/s (P = .022).

When compared to 90 days and to baseline, the nutritional status worsened at 180 days (Table 2). Three NSPRG participants who were at undernutrition risk at 90 days become undernourished at 180 days (Table 2). The proportion of CG patients at risk of undernutrition increased from 34.9% at baseline to 58.1% at 180-day follow-up. No significant changes were found for functional status in this period. Regarding cognitive status, compared to 180 days from baseline, a slight decrease was noticed in the NSG and in the NSPRG of −1.1 (3.2) and −1.7 (5.1) points, respectively (P < .05; Table 3).

General Linear Model Analysis

The NSG ▵MNA score after intervention, at 21 and 90 days of follow-up, is independent of the MNA and the MMSE scores at baseline (Ps > .05). At 180 days of follow-up, the NSG ▵MNA score is dependent on the MNA score at baseline (P = .012; adjusted R² = .102). In the NSPRG, the ▵MNA score after intervention, at 21, 90, and 180 days of follow-up, is independent of the MNA and MMSE scores at baseline (Ps > .05; Table 5).

Discussion

In the present study, a 21-day small volume high-protein energy-dense oral liquid supplement and a psychomotor rehabilitation program had a positive effect on nutritional and functional status of community-dwelling older adults with probable mild AD. An increase in MNA scores in both intervention groups (NSG and NSPRG) and body weight (kg) in the NSPRG was observed. Moreover, these higher MNA scores were maintained after discontinuing intervention at 90-day follow-up. Another positive finding was the functional status improvement among NSPRG at 90 days of follow-up. Nevertheless, at 180 days of follow-up, a slight decrease for other studied nutritional parameters such as MUAMC and FFM (%) was seen in both intervention groups (NSG and NSPRG), although continuation in functionality and autonomy was observed in the entire sample. It is important to highlight that at 90 days of follow-up, the NSPRG maintains the DMNA▵ score improvements independent of the MNA score and MMSE score at baseline.

According to our knowledge, there are no previous intervention studies using both ONS and psychomotor rehabilitation program to which the present results can be compared. However, reports from intervention with oral supplements are available. Faxén-Irving et al ²⁰ showed that a 6-month ONS (400 mL, 410 kcal, 18 g protein) in elderly patients with dementia in day care improved their body weight, although no significant changes were found in the ADLs. Lauque et al, ¹⁷ after a 3-month ONS (300-500 kcal, 10-20 g protein) intervention in patients with AD on geriatric wards and day care centers, reported an increase in body weight, FFM, and MNA score but not on the functional and cognitive status. Similarly, Planas et al ²¹ showed that patients with mild AD in day care centers receiving 6-month, twice-a-day 250 mL energy-dense and protein-rich ONS 500 kcal/d (45% carbohydrate, 25% fat, 30% proteins and micronutrients) increased BMI, TST, and MUAMC, but no improvements for functional and cognitive status were reported. In addition, Pivi et al, ¹⁹ in a 6-month randomized controlled trial, controlled with a nutrition education group, found that a twice-a-day ONS, providing 680 kcal and 25.6 g protein/day, is more effective to improve the nutritional status of community-dwelling patients with AD, increasing the body weight, BMI, mid-upper arm circumference, and MUAMC. Despite this, none of these previous studies demonstrated an effect of the interventions on physical performance (muscle function) measured by GS.

Considering that weight loss ⁷ and gait dysfunction ²⁴ have been recognized as clinical markers of AD, ²⁵ which in turn increase the loss of independence and safe mobility,²⁶ it is important to highlight the MNA score and GS improvements in the NSPRG found after intervention at 21 and 90 days, respectively, and also the MNA score maintenance after discontinuing intervention at 90-day follow-up. These findings suggest a positive effect between nutritional and functional status, suggesting that the ONS and psychomotor rehabilitation program intervention is beneficial to ameliorate the balance and gait impairments associated with AD. ^25,26

In a systematic review, Abellan van Kan et al ⁴⁰ concluded that GS measured at usual pace could be considered a strong and consistent predictor of adverse outcomes in community-dwelling older people. ⁴⁰ Furthermore, Gras et al ²⁶ recently reported that community-dwelling older adults with very mild AD and gait deficits benefited from an early physical therapy slowing the progression of further balance and gait dysfunction. These findings may suggest that more specific rehabilitation programs are required for functional improvements to become noticeable.

It is therefore debatable whether the intervention focused solely on SDA and ONS is enough to improve the nutritional and functional status. Indeed, in a recent systematic review by Liu et al, ⁴⁸ it was reported that “nutritional supplements,” “training/education programs,” “environment/routine modification,” or “feeding assistance” showed low to moderate evidence in improving food intake, body weight, and BMI in older adults with dementia.

Evidence is sparse concerning positive changes of nutritional, functional, and cognitive status in community-dwelling patients with AD by any interventions. However, considering that decline in body composition and physical performance is associated with AD progression ⁴⁹ as well as with a negative impact on health status and quality of life, ^6,10 present results reveal the importance of providing an early ONS and a psychomotor rehabilitation program intervention in undernourished or nutritionally at-risk community-dwelling older adults with mild probable AD. Besides, our results reinforce the importance of multidisciplinary intervention in community-dwelling older adults with mild probable AD.

Regarding the cognitive function, a small decrease in the MMSE score was observed in both intervention groups (NSG and NSPRG) on the 180th day of follow-up. Though, in a recent review of randomized controlled trials, it was shown that aerobic exercises ameliorate cognitive function in older adults with mild cognitive impairment. ⁵⁰

A strength of the present study was the high compliance to the ONS and the adherence to the psychomotor rehabilitation program by the patients, family, caregivers, and the nursing staff. As expected and according to the previously described by Lombard et al, ⁵¹ a small volume high-protein energy-dense liquid ONS (125 mL) has shown to improve compliance in clinical trials, optimizing the clinical effectiveness of oral nutritional supplements. A limitation of our study that should be acknowledge is the lack of a random sample allocation of the NSPRG participants. Moreover, patients’ selection was made according to their baseline MMSE score and affinity and this situation may limit the comparison between both intervention groups, the NSG and the NSPRG. It should also be acknowledged that psychiatric diagnosis was confined to AD and the possible confounding effect of depression or of subsyndromal depressive symptoms was not considered in the present analyses.

Increasing evidence showing that nutritional factors can influence both the risk of developing AD and its rate of clinical progression ³ suggests that the association between diet, nutritional status, and cognitive function deserves more attention. ³ Though, considering that older adults in early stages of AD may show weight loss and gait dysfunction, ²⁶ the present study results could have an important implication in clinical procedures, suggesting that nutritional approach in undernourished or nutritionally at-risk community-dwelling older adults with probable mild AD needs to be more targeted to physical performance (muscle function) and consequently to nutritional and cognitive status.

In conclusion, the present study shows that a 21-day small volume high-protein energy-dense ONS and a psychomotor rehabilitation program improve long-term nutritional and functional status of community-dwelling older adults with probable mild AD. However, large-scale randomized controlled trials focused on the nutritional intervention and motor rehabilitation in community-dwelling mild AD are required in order to clarify the beneficial effects.