The LIBRA Index in Relation to Cognitive Function,Functional Independence,and Psycho-Behavioral Symptoms in a Sample of Non-Institutionalized Seniors at Risk of Dementia

Abstract

Background:

Alzheimer’s disease is the principal cause of dementia and is determined, in at least one third cases, by modifiable risk factors (MRF). The “Lifestyle for Brain Health (LIBRA)” index was recently developed to quantify the individual risk of progression to dementia ascribable to MRF.

Objective:

The aim of this study was to investigate the association between LIBRA scores and markers of cognitive performance, functional independence, and psycho-behavioral symptoms in a community-based sample of Italian elders.

Methods:

308 senior participants with mild cognitive impairment (MCI) or subjective cognitive decline (SCD) were evaluated with a complete neuropsychological battery and semi-structured interviews for the assessment of depression, apathy, and functional autonomy. All the 12 LIBRA MRF were available for the calculation of LIBRA scores. A modified version of the index (LIBRA-2) was calculated by removing depression weight from the LIBRA index. Partial correlation analyses, controlling for age and education, assessed the association between LIBRA indices and cognitive, functional, and behavioral outcomes. Separate analyses were repeated in the MCI and SCD subgroups.

Results:

In participants with SCD (SCDp), significant correlations existed between LIBRA and markers of impairment in global cognition, visuo-spatial attention, and semantic fluency. LIBRA-2 associated with psycho-behavioral symptoms in the whole sample and in SCDp. LIBRA-2 only associated with apathy in the MCI subgroup.

Conclusions:

The LIBRA index might be useful to determine the lifestyle-attributable risk of cognitive and psycho-behavioral decline in Italian seniors at risk, while in those with overt cognitive impairment, these outcomes are presumably mainly associated with non-modifiable factors.

Keywords

Alzheimer’s disease apathy depression LIBRA Index modifiable risk factors mild cognitive impairment neuropsychological tests subjective cognitive decline

INTRODUCTION

The number of people affected by Alzheimer’s disease (AD) is expected to double every 20 years, due to the progressive aging of the population, reaching nearly 115 million in 2050 [1, 2]. This calculation is based on stable dementia prevalence rates, stratified by age [3]. However, some evidence suggest that the age-specific incidence is now decreasing, at least in high income countries, presumably due to an improvement in lifestyle conditions, higher educational levels, and better management of modifiable risk factors (MRF) for dementia [4, 5]. Indeed, according to a review of 2011, 7 main MRF—diabetes, hypertension, obesity, low education, low physical activity, smoking, and depression—account for a relevant proportion of AD cases [6]. Interventions based on these MRF have been estimated to reduce the incidence of AD by 10–25% so as to prevent up to three million new AD cases in 2050 worldwide [7].

People affected by AD have a long asymptomatic phase prior to dementia onset. With the progression of the disease, individuals with AD may show a mild cognitive decline which may be accompanied by subtle deficits in instrumental activities of daily living (IADL). This transitional zone between normal aging and dementia is called mild cognitive impairment (MCI) [8]. Although not all people with MCI (MCIp) progress to dementia over time, with some MCIp reverting spontaneously to normal cognition [9], MCI is considered a high-risk condition for dementia and MCIp are rated the best candidates for potential treatments aimed at preventing or postponing its onset [10]. MCI may, in turn, be preceded by a phase in which the individual reports the self-perceived presence of cognitive difficulties that are hard to detect objectively, since his/her performance in neuropsychological evaluations falls within the ranges of normality (subjective cognitive decline, SCD) [11]. Similarly to MCIp, SCDp are also at increased risk of progression to AD dementia [12, 13]. In particular, executive and language SCD complaints seem to discriminate preclinical AD from normal aging [14].

Available pharmacological treatments for AD are ineffective in modifying the course of the disease [15]. Thus, attention has progressively focused on the promotion of healthy lifestyles and on the evaluation of preventive programs and interventions targeting MRF for dementia [16].

For the purpose of an early identification of persons at risk and a better quantification of the individual risk of progression, several indices have been recently proposed, including the Cardiovascular Risk Factors, Aging and Incident of Dementia (CAIDE) score [17] and the Australian National University Alzheimer’s Disease (ANU-ADRI) index [18, 19]. Those indices demonstrated useful in detecting persons who will convert to dementia many years before the onset of the disease, however, since they base also on non-modifiable risk factors (n-MRF), such as age, sex, and Apolipoprotein E (APOE) genotype, they are of limited utility in addressing preventive programs.

The LIfestyle for BRAin Health (LIBRA) index was recently developed by researchers from Maastricht University, in order to include only modifiable dementia risk factors, which can be targeted by lifestyle interventions or primary prevention strategies [20]. The LIBRA index includes 12 MRF, i.e., depression, diabetes, cognitive activity, physical inactivity, hypertension, Mediterranean diet, obesity, smoking, low/moderate alcohol consumption, hypercholesterolemia, coronary heart disease, and renal dysfunction, that were selected according to expert opinion and empirical evidence from systematic review of observational epidemiological studies, and then weighted, in order to obtain a balanced account of their relative contribution to dementia risk (see Deckers et al. [21] for meta-analyses), by dividing the natural logarithms (ln) of their relative risk (RR) by the lowest ln (RR) (i.e., by 0.30 for low/moderate alcohol consumption) [20, 21]. For each individual, a LIBRA score can be calculated simply by summing the partial scores of each MRF to which he/she is exposed [21]. The LIBRA index was validated on a large longitudinal population-based sample, which was followed up for 16 years [20]. The authors observed that one-point increase in LIBRA scores resulted in an augmented risk of dementia of 19% and in a 9% increased risk of cognitive decline.

A direct association between LIBRA scores and the extent of dementia risk was also observed in middle-aged persons and elders between 55 and 79 years, while in persons aged 80 or more years, the LIBRA index seemed not useful in predicting progression to dementia [22, 23].

At present, no study has investigated the LIBRA index as a potential tool to identify people at risk of dementia in the Italian population. The aims of this study are therefore:

To calculate the LIBRA index (i.e., to evaluate the presence and the number of MRF) in a non-clinical sample of Italian older individuals at risk of developing dementia.

To verify the presence of any association between LIBRA scores and scores from an extensive battery of neuropsychological tests, psycho-behavioral scales, and functional indices.

To inquire if LIBRA scores and their association with cognitive, functional, or behavioral outcomes differ between MCIp and SCDp.

METHODS

Participants

A cohort of older individuals at risk, aged between 60 and 90, was recruited among those attending Senior Centers in Rome (Italy). Only seniors who were willing to be enrolled in a randomized controlled trial (GR-2013-02356043, co-financed by the Italian Ministry of Health) aimed to assess the effectiveness of a 12-week intervention of cognitive stimulation and/or physical exercise in preventing dementia or cognitive and functional decline were included. Data of this cross-sectional analysis are those pertaining to the baseline evaluation of the participants.

Inclusion criteria were: age ≥60 years; Mini-Mental State Examination (MMSE) score [24] ≥20; subjective cognitive decline (SCD) in one or more cognitive domains and/or objective cognitive impairment in at least one test from a complete neuropsychological battery (specified in Neuropsychological Assessment below); absence of a significant functional impairment, that was operationalized as a score <9 in the Functional Assessment Questionnaire (FAQ) [25] or as a loss <20% of functionality in the Instrumental Abilities of Daily Living (IADL) [26]. Exclusion criteria were: MMSE score <20 or significant functional impairment according to FAQ and IADL scales; diagnosis of dementia or other neurologic or clinical conditions that could affect cognitive performances; history of severe acquired brain injury; major psychiatric disease; alcohol or substance dependence; and inability to provide informed written consent.

The study was approved by the Ethical committee of “Fondazione Santa Lucia IRCCS” and conducted according to international standard of Good Clinical Practice (GCP).

Clinical assessment

Information about demographics, lifestyle habits (caffeine, alcohol, and tobacco consumption), subjective cognitive complaints, and clinical data (pharmacological treatment, presence of diabetes, hypertension, hypercholesterolemia, thyroid dysfunction, cardiovascular disorders, cancer, hospitalization or access to emergency department in the year preceding the assessment) were collected through a standardized interview, that was administered by trained researchers. Screening questions for psychotic symptoms of the Structured Clinical Interview 1 (SCID-I) were also administered, in order to exclude people with significant clinical signs of psychosis.

Neuropsychological assessment

A complete neuropsychological battery was administered by trained neuropsychologists, including the MMSE [24] as a screening test for evaluation of global cognition, and domain-specific scales [27 –34]: the Digit span forward (DF) and backward (DB) [27] were administered for the assessment of short-term memory and working memory/attention, respectively. DF requires the subject to repeat numerical sequences of increasing length in the same order as read aloud by the examiner, while in DB the individual has to repeat numbers in the reverse order of that presented. In each test, scores correspond to the maximum number of digits that the examinee is able to repeat correctly.

Immediate and delayed recall of the Short Story Test (SS IR and SS DR) [28] were used to evaluate episodic verbal memory. In this test, the examiner reads a short story aloud, which the subject must repeat immediately and after a 10-minute delay. In each test, the participant’s score corresponds to the maximum number of elements that he/she is able to report correctly.

The Line Cancellation Test (LCT) [29] and the Multiple Feature Target Cancellation (MFTC) [30] were adopted to assess efficiency and speed of visual scanning and attention, together with the ability to deal with multiple occurrences of the same target across a page, while inhibiting distractors and targets that have already been marked. In LCT examinees must cross out 60 lines that are placed in random orientations on a paper sheet; MFTC requires to identify and circle 13 target items that are presented in an array among distractors. In both tests, the time required for completion (in seconds) and the number of identified targets are registered; in the MFTC, the accuracy in visual search is also evaluated with a mathematical formula that combines correct hits and false alarms [30].

Constructional praxis was investigated with the Rey-Osterrieth Complex Figure Test (ROCFT) [31], in which the participant is asked to reproduce by hand a complicated abstract line drawing composed of 18 main elements: 2 points are assigned to each correctly copied and positioned element, 1 point to each wrongly positioned or wrongly reproduced (yet recognizable) element, 0.5 to each a wrongly positioned and wrongly reproduced element. ROCFT scores therefore may vary between 0 and 36.

The Frontal Assessment Battery (FAB) was used for the assessment of executive functions [32]. The FAB contains 5 sub-tests, investigating conceptualization, cognitive flexibility, motor planning, sensitivity to interference, inhibitory control, perseveration, and environmental autonomy. A maximum score of 3 is assigned to each sub-test, for a maximum overall raw score of 18.

Verbal fluency and naming abilities were assessed with the phonemic and semantic fluency tests (PF and SF) and with the Naming from Description test (ND). In PF and SF [33], the examinee has 60 seconds to produce as many words as possible that begin with a given letter (i.e., “S”) or pertain to a semantic category (i.e., “Animals”), except for proper names, numbers, and sets of words sharing the same lexical root. In both tests, scores correspond to the maximum number of word that the examinee is able to name, in compliance with the rules. In ND, a verbal description of each of 38 words, belonging to the categories of concrete living/non-living and abstract concepts, is read aloud by the examiner. Participants are instructed to listen each definition and to provide the name of the described concept. For each item, the examinee obtains a score of 1 if he/she autonomously provides the correct answer, a score of 0.5 if he/she is able to identify it among 3 alternatives and a score of 0 if no/wrong answer is provided. The overall score is calculated by summing the scores per-item [34].

Functional assessment

As previously reported, levels of independence in IADL were assessed with the Functional Assessment Questionnaire (FAQ) [25], which measures participants’ ability to manage 10 activities of daily living, such as preparing balanced meals, keeping track of current events, managing money and documents, and others. Each activity is assigned a score ranging from 0 to 3, higher scores correspond to greater functional dependence. The overall FAQ score, which is obtained by summing the partial scores, ranges from 0, indicating complete independence in IADL, to 30, corresponding to complete impairment. A cut-point of 9 (dependent in 3 or more activities) is recommended to define the presence of clinically relevant functional impairment [35, 36]. Therefore, as previously reported, individuals scoring 9 or more were excluded from the study. Only disabilities that were consequent/consistent with the cognitive difficulties were considered for evaluation, while functional impairment related to pre-existing sensory-motor problems or not due to cognitive impairment was excluded from the calculation of FAQ scores.

Physical activity

Physical activity was assessed with the “International Physical Activity Questionnaire – short form” (IPAQ-SF) [37]. IPAQ-SF records the individual’s activity according to four intensity levels: 1) vigorous-intensity activity, such as aerobics, 2) moderate-intensity activity, such as leisure cycling, 3) walking, and 4) sitting. The time spent in each activity level can be converted into Metabolic Equivalent of Task (MET) values, in order to obtain an index of the amount of the individual’s total energy expenditure.

Mediterranean diet

The “Mediterranean Diet Adherence Screener” (MEDAS) [38, 39] was administered to investigate food habits.

The MEDAS is a 14-item questionnaire requesting participants to report food habits (consumption of olive oil and greater consumption of white meat, compared to red meat) and frequency of consumption/amount of 12 main foods related to the Mediterranean diet. Each of the 14 MEDAS items is assigned a score of 0 or 1 point, according to pre-determined criteria of adherence to the Mediterranean diet [38, 39]. The maximum overall score is 14.

Cognitive activity

A self-report questionnaire was created by the authors to investigate participation in 16 cognitively stimulating leisure activities or hobbies. Each activity/hobby was assigned a score of 0 or 1, according to pre-defined frequency cut-offs (i.e., “at least once a week” for reading, following economic, social, politic, or other news/events, writing, engaging in arts, playing mind games like crosswords, cards, Sudoku, and others, knitting or embroidery, cooking, doing housework, gardening, attending senior centers or other social groups, hanging out with friends, performing other indoor or outdoor leisure activities; “at least once a month” for going to the theater/cinema, attending courses, volunteering and charitable activities; “at least every 6 months” for travelling). The overall score was calculated by summing partial scores and ranged from 0 to 16.

Psycho-behavioral symptoms

The Geriatric Depression Scale – 5-item (GDS-5) [40], the Apathy Evaluation Scale (AES) [41], and the screening questions for psychotic symptoms of the Structured Clinical Interview 1 (SCID-I) were administered in order to screen for depressive, apathetic, and psychotic symptoms, respectively. As previously reported, individuals with significant psychotic conditions were excluded, while depression and apathy evaluations were considered psycho-behavioral outcomes for this research.

The GDS-5 is a short questionnaire investigating life satisfaction, social withdrawal, feelings of emptiness/boredom, helplessness, and worthlessness. A point is assigned to the presence of each of these 5 items, resulting in a global GDS-5 score ranging from 0 to 5. The 18 questions of AES inquire 3 domains of apathy (decreased goal-directed behaviors, reduction of goal-related thoughts, emotional indifference with flat affect). Each item is scored on a 4-point Likert Scale, with overall AES scores ranging from 18 to 72. Higher scores reflect more severe apathy.

LIBRA index

Information was available on 12 out of 12 of the risk/protective factors of LIBRA [21]. The criteria used to assign scores associated to each MRF are described below and summarized in Table 1.

Table 1

Criteria used to calculate the LIBRA scores from the assessment of the 12 LIBRA factors

Factor	Criteria	Score
Modifiable risk/protective factors
Depression	GDS-5 ≥2	+2.1
	GDS-5 <2	0
Hypertension	Yes	+1.6
	No	0
Hypercholesterolemia	Yes	+1.4
	No	0
Diabetes	Yes	+1.3
	No	0
Renal dysfunctions	Yes	+1.1
	No	0
Coronary heart disease	Yes	+1.0
	No	0
Obesity	BMI ≥30	+1.6
	BMI <30	0
Mediterranean Diet	MEDAS <9	0
	MEDAS ≥9	–1.7
Smoking	Yes	+1.5
	No	0
Physical activity	IPAQ-SF <600 MET	+1.1
	IPAQ-SF ≥600 MET	0
High cognitive activity	Leisure activities/hobbies ≥8	–3.2
	Leisure activities/hobbies <8	0
Low/moderate alcohol use	0 < drinks/day <2 or 0 < drinks/week <14	–1.0
	0 drinks/day or ≥2 drinks/day or ≥14 drinks/week	0

GDS, Geriatric Depression Scale; BMI, body mass index; MEDAS, Mediterranean Diet Adherence Screener; MET, metabolic equivalent of task; IPAQ-SF, International Physical Activity Questionnaire Short Form.

Depression. A GDS-5 score ≥2 was considered as indicative of the presence of clinically significant depression.

Diabetes. Participants were considered as diabetics in the presence of both diagnosis of diabetes mellitus (DM) in medical records or self-reported type 2 DM during clinical interview, and documented use of oral hypoglycemic drugs and/or insulin therapy.

Physical inactivity. Physical activity was measured by converting activity bouts in MET values, according to the IPAQ-SF scoring guidelines [37]. A cut-point of 600 MET/week, corresponding to a moderate level of physical activity in IPAQ-SF, was adopted to classify participants as physically active or inactive.

Hypertension. Limited evidence exists on the potential role of late-life hypertension as a risk factor for dementia [42]. However, similarly to Schiepers et al. [21], participants with anamnestic report of hypertension or using antihypertensive medications were classified as hypertensive, regardless of age of onset.

Obesity. There is not univocal evidence regarding the role of obesity in late-life as a risk factor for dementia [20]. However, similarly to Schiepers et al. [21] and Pons et al. [23], persons with a body mass index (BMI) ≥30 were classified as obese, according to the WHO definition [43], independently from age of onset.

Smoking. Participants were screened for cigarette smoking and classified as non-smokers if they responded “no” and as smokers if they responded “yes” to the question: “Do you currently smoke?”. Detailed information about the number of cigarettes/ tobacco products consumed per day and previous smoking habits in ex-smokers was also collected, yet not used in the calculation.

Hypercholesterolemia. A recent meta-analysis of five prospective studies reported a 54% increased risk of dementia in older adults with high levels of serum cholesterol [18]. Therefore, similar to Pons et al. 2017 [23], participants with anamnestic report of hypercholesterolemia or using statins or other medications for hypercholesterolemia were rated as positive, independently from age of onset.

Coronary heart disease. Participants with anamnestic report of ‘previous or current coronary heart disease’ and documented use of anti-angina medications, daily antiarrhythmic drugs or by history of coronary artery revascularization procedures were classified as positive.

Renal dysfunction. Participants with anamnestic report of any kind of symptomatic renal problems and documented use of medications for chronic kidney dysfunction or dialysis were rated as risk factor positive.

High cognitive activity. A cut-score of 7 in the self-report questionnaire about participation in cognitively stimulating leisure activities/hobbies (corresponding to the highest tertile of the sample) was adopted to classify participants as cognitively active or inactive.

Low to moderate alcohol consumption. Based on the National Health and Medical Research Council (NHMRC) guidelines [44], low/moderate alcohol intake was rated as 0 < standard drinks per day < 2 or as < 0 standard drinks per week < 14. Being a teetotaler was not rated as a protective factor.

Mediterranean diet. Participants who obtained a MEDAS score <9 were classified as non-adherent to the Mediterranean diet and assigned a LIBRA score of 0 [45 –49].

Statistical analyses

Statistical analyses were performed using the software IBM SPSS Statistic, version 21.0 (IBM Statistics, IBM Corporation, New York, NY, USA). Participants characteristics were presented as absolute frequencies and percentages for categorical variables and as means±standard deviations (SD) for continuous variables.

Pearson’s correlation test was used for the analysis of the association between LIBRA scores and scores from neuropsychological, functional and psycho-behavioral evaluation scales.

Since depression represents one of the 12 LIBRA risk factors, an 11-item index, the LIBRA-2, was computed by removing the depression weighted score from the LIBRA index. Correlation analysis was then repeated between LIBRA-2 and GDS-5 and AES scores, with the aim to verify if the remaining 11 MRF associated with depressive and apathetic symptoms.

Since age and education have a demonstrated influence on cognitive performance, partial correlation analyses were performed, in order to determine the existence of any relationship between LIBRA and study outcomes, and between LIBRA-2 and scales for apathy and depression, whilst controlling for age and schooling.

The neuropsychological scores of each participant were then standardized, according to Italian normative values [24 , 27–34]. Participants obtaining a standardized (Z) score > –2 in all neuropsychological tests were included in the “Subjective Cognitive Decline (SCD)” group, while seniors who obtained a Z score ≤–2 in at least one neuropsychological test were attributed to the “Mild Cognitive Impairment (MCI)” group.

Then, all the scores obtained by the SCD and MCI groups were compared with crude analyses (T-Test for independent sample) and Analyses of Variance (ANOVAs), adjusted for gender, age, and education.

Bivariate correlation and partial correlation analyses were then repeated in SCD and MCI subgroups, in order to verify if a possible differential association between LIBRA scores and study outcomes existed in individuals with and without initial cognitive decline.

RESULTS

Descriptives

A total of 327 elders were screened for inclusion (mean age 74.07±6.20; mean education 9.54±4.23; 25.30% men). Six persons (1.82%) were excluded due to significant functional impairment in FAQ or IADL, 7 (2.13%) for concomitant neurologic or clinical conditions that could affect cognitive performances, 2 (0.60%) for major psychiatric disease, and 2 (0.60%) for significant physical/motor impairment. Three people (0.91%) refused to complete the entire baseline evaluation. Complete data on all 12 LIBRA MRF were obtained from 308 (93.90%) participants (mean age 73.40±6.30; mean instruction 9.30±4.09; 25.6% males) who were included in subsequent analyses.

Participants obtained a mean LIBRA score of 1.59±2.61 (median = 1.60; min = –4.20; max =8.80); the mean MMSE score was 27.70±1.90 (median = 28.00; min = 22.00; max = 30.00); mean FAQ was 1.58±2.02 (median = 1.00; min = 0.00; max = 9.00).

Half of the participants (N = 154; 50.0%) met criteria for MCI, while the remainders were classified as having a SCD.

Detailed data about demographics and scores from the cognitive, functional, and psycho-behavioral evaluations are reported in Table 2.

Table 2

Demographics and standardized scores from the cognitive, functional, psycho-behavioral evaluations in the whole sample and in MCI and SCD subgroups (MCIp and SCDp). Unless otherwise indicated, the results of the statistical comparisons adjusted for demographics are reported

			Subgroup
			SCDp (n = 154)	MCIp (n = 154)	Total (n = 308)	F	p
Sex	(females)	Count (%)	114 (74%)	115 (74.7%)	229 (74.4%)	0.17°	0.896
Age		Mean±SD	72.77±5.95	75.06±6.47	73.91±6.31	–3.24°°	0.001
Education			11.41±3.96	8.13±3.84	9.76±4.22	7.38°°	<0.001
LIBRA			1.75±2.50	1.43±2.71	1.59±2.61	0.331	0.566
MMSE			28.51±1.32	26.88±2.02	27.69±1.89	24.10	<0.001
DF			–0.04±1.04	–0.49±1.15	–0.26±1.12	5.46	0.020
DB			–0.01±0.99	–1.06±1.49	–0.54±1.37	20.65	<0.001
SS IR			0.11±1.15	–1.09±1.17	–0.49±1.30	39.41	<0.001
SS DR			0.03±1.13	–1.08±1.14	–0.53±1.26	37.14	<0.001
LCT t (s)			0.51±0.37	0.33±0.50	0.42±0.44	1.23	0.267
LCT a			0.27±0.73	–0.99±3.29	–0.35±2.45	10.13	0.002
MFTC t (s)			1.11±0.37	0.93±0.56	1.02±0.48	2.80	0.095
MFTC a			–0.13±0.91	–1.56±1.78	–0.83±1.57	30.29	<0.001
FAB			0.13±0.95	–0.84±1.31	–0.36±1.24	23.07	<0.001
PF			0.97±1.15	–0.02±1.11	0.47±1.23	8.99	0.003
SF			0.30±1.00	–0.30±0.92	0.00±1.00	22.42	<0.001
ND			0.33±0.64	–0.86±1.54	–0.26±1.32	27.77	<0.001
ROCFT			0.07±0.76	–1.20±1.51	–0.56±1.35	35.22	<0.001
FAQ			1.37±1.85	1.81±2.16	1.59±2.02	0.561	0.454
GDS-5			1.18±1.43	1.24±1.42	1.21±1.42	0.613	0.434
AES			29.22±6.25	29.87±6.27	29.54±6.26	0.327	0.568

Domain-specific Neuropsychological tests scores are standardized according to means and standard deviations of normative Italian groups, in order to allow for an easier readability and comparability between the different tests. MCIp, participants with mild cognitive impairment; SCDp, participants with subjective cognitive decline; DF, digit span forward; DB, digit span backward; SS IR, Short Story Immediate Recall: SS DR, Short Story Delayed Recall; LCT t (s), Line Cancellation Test time (s); LCT a, Line Cancellation Test accuracy; MFTC t (s), Multiple Features Target Cancellation time (s); MFTC a, Multiple Features target cancellation accuracy; FAB, Frontal Assessment Battery; PF, phonological fluency; SF, semantic fluency; ND, naming from description; ROCFT, Rey-Osterrieth Complex Figure Test; FAQ, Functional Assessment Questionnaire; GDS-5, Geriatric Depression Scale - 5 items; AES, Apathy Evaluation Scale. ° Results of the Chi-2 test with 2 degrees of freedom (d.f.). °° Results of T-tests for independent samples with 306 d.f.

Whole sample analyses

In the whole sample, a mild inverse correlation was observed between LIBRA and MMSE scores (Pearson’s R = –0.15; p = 0.007). LIBRA was also weakly associated to SF (R = –0.16; p = 0.004) and PF (R = –0.12; p = 0.034) (Table 3). All the correlations between the LIBRA score and scores from domain specific neuropsychological tests and functional scales are shown in Supplementary Table 1. LIBRA showed also a moderate direct correlation with GDS-5 (R = 0.46; p < 0.01) and AES (R = 0.33; p < 0.01) (Table 3).

Table 3

Results of correlation analyses between LIBRA and LIBRA-2 score and demographics, cognitive, functional and psycho-behavioral test scores in the whole sample of 308 participants. Significant associations (p < 0.05) are reported in bold

		LIBRA		LIBRA-2
		(n = 308)		(n = 308)
		Raw	Adjusted	Raw	Adjusted
Age	R	0.071	–	0.041	–
	p	0.211	–	0.470	–
Education	R	–0.128	–	–0.121	–
	p	0.025	–	0.034	–
FAQ	R	0.098	0.084	0.068	0.056
	p	0.084	0.142	0.236	0.327
MMSE	R	–0.153	–0.096	–0.156	–0.111
	p	0.007	0.094	0.006	0.053
PF	R	–0.121	–0.063	–0.121	–0.072
	P	0.034	0.275	0.035	0.209
SF	R	–0.162	–0.112	–0.146	–0.104
	p	0.004	0.051	0.010	0.071
GDS-5	R	0.468	0.461	0.274	0.264
	p	<0.001	<0.001	<0.001	<0.001
AES	R	0.334	0.319	0.268	0.256
	p	<0.001	<0.001	<0.001	<0.001

LIBRA, Lifestyle for Brain Health; LIBRA-2, Lifestyle for Brain Health after removing depression weight; FAQ, Functional Assessment Questionnaire; MMSE, Mini-Mental State Examination; PF, phonological fluency; SF, semantic fluency; GDS-5, Geriatric Depression Scale - 5 items; AES, Apathy Evaluation Scale; Raw: results of the bivariate correlation analysis; Adjusted: results of the partial correlation analysis controlling for age and education.

Table 4

Results of correlation analyses between LIBRA and cognitive, functional and psycho-behavioral test scores in subgroups of participants with MCI and SCD. Significant associations (p < 0.05) are reported in bold

		MCIp		SCDp
		(n = 154)		(n = 154)
		Raw	Adjusted	Raw	Adjusted
MMSE	R	–0.061	0.005	–0.262	–0.237
	p	0.451	0.947	0.001	0.003
LCT t (s)	R	–0.109	–0.159	0.237	0.202
	p	0.225	0.078	0.007	0.023
SF	R	–0.082	–0.016	–0.210	–0.181
	p	0.312	0.849	0.009	0.026
ROCFT	R	0.054	0.131	–0.184	–0.151
	p	0.506	0.109	0.023	0.064
FAQ	R	0.037	0.027	0.154	0.147
	p	0.649	0.740	0.057	0.071
GDS-5-L2	R	0.181	0.158	0.362	0.358
	p	0.025	0.053	<0.001	<0.001
AES-L2	R	0.198	0.187	0.332	0.322
	p	0.015	0.022	<0.001	<0.001

MCIp, participants with mild cognitive impairment; SCDp, participants with subjective cognitive decline; MMSE, Mini-Mental State Examination; LCT t (s), Line Cancellation Test time (s); ROCFT, Rey-Osterrieth Complex Figure Test; FAQ, Functional Assessment Questionnaire; GDS-5-L2, Geriatric Depression Scale - 5 items - correlations with LIBRA-2.; AES-L2, Apathy Evaluation Scale - correlations with LIBRA-2.; R, Pearson’s R; Raw: results of the bivariate correlation analysis; Adjusted: results of the partial correlation analysis controlling for age and education.

In partial correlation analyses, controlling for age and education, LIBRA was significantly associated with GDS-5 (R = 0.46 p < 0.01) and AES (R = 0.31 p < 0.01); no significant correlation was observed between LIBRA and cognitive or functional outcomes.

After removing the depression weights from the LIBRA score calculation, a mild association between LIBRA-2 and GDS-5 (R = 0.26 p < 0.01) and AES (R = 0.25 p < 0.01) was still detectable.

Separate analyses for SCDp and MCIp subgroups

SCDp were significantly younger (T-test, 306 df = –3.24; p = 0.001) and more educated (T-test, 306 df = 7.38; p < 0.01) than MCIp. As expected, SCDp performed significantly better than MCIp in all neuropsychological tests (Table 2). No significant differences between subgroups were observed in relation to functional and psycho-behavioral indicators, even if MCIp obtained slightly higher FAQ, AES, and GDS-5 scores than SCDp (Table 2). The LIBRA scores of the MCIp were slightly and not significantly higher than those of the SCDp (T-test, 306 df = –1.08; p = 0.281; Table 2 reports the results of the statistical comparison after adjustment for demographic variables).

In SCDp LIBRA scores correlated weakly with MMSE scores (R = –0.26; p = 0.001), LCT execution times (R = 0.23; p = 0.007), SF scores (R = –0.21; p = 0.009), and ROCFT scores (R = –0.18; p = 0.023) (Table 4) while a moderate correlation was observed between LIBRA-2 (R = 0.36; p≤0.001) and GDS-5 and AES scores (R = 0.33; p≤0.001). When controlling for demographics, a mild correlation was observed between LIBRA and MMSE (R = –0.23; p = 0.003), LCT execution times (R = 0.20; p = 0.023), and SF scores (R = –0.18; p = 0.026), while LIBRA-2 was moderately associated with GDS-5 (R = 0.35; p < 0.001) and AES-5 (R = 0.32; p < 0.001).

In the MCIp subgroup, no correlation was observed between LIBRA scores and neuropsychological scores, while a mild direct association was observed between LIBRA-2 and GDS-5 (R = 0.18; p = 0.025) and AES scores (R = 0.20; p = 0.015) (Table 4). When controlling for demographics, a mild correlation was observed between LIBRA-2 and AES (R = 0.19; p = 0.022).

DISCUSSION

In the literature, several indices for dementia risk have been proposed to quantify the impact of modifiable and non-modifiable predictors of progression to dementia and to identify persons at higher risks of progression. All, except LIBRA, include demographic and n-MRF: for example, the Cardiovascular Risk Factors Aging and Dementia (CAIDE) [17], which takes account of age, sex, education, systolic blood pressure, total cholesterol, obesity, physical inactivity, and presence of the APOE ɛ4 allele, or the Australian National University Alzheimer’s Disease Risk Index (ANU-ADRI) [18] which includes 11 risk and 4 protective factors, among n-MRF and MRF. The ANU-ADRI was validated on three independent cohorts of older adults in 2014 [19]. A self-administered online version of the questionnaire is also available online. The inclusion of the n-MRF increases the predictive value of these indices—for example, CAIDE demonstrated useful in predicting dementia risk in the following 20 years [17]—yet, at the same time, it makes them less feasible and hardly suitable for prevention approaches. Indeed, the presence of n-MRF could be confounding and could lead to a possible underestimation of the efficacy of interventions.

The LIBRA index is a measure of the individual risk to progress to dementia, that is based only on modifiable risk or protective factors. Its strengths reside in being suitable both for lifestyle multidomain interventions, independent from the subject’s life habits, and for personalized approaches targeting individualized MRF for dementia.

The association between LIBRA scores and cognitive performances has already been evaluated by other studies, which are briefly reviewed below. However, each of them lacked one or more factors that were necessary for the calculation of LIBRA scores, while this is the first research study including all 12 LIBRA MRF.

On a large multicentric European population-based cohort [21] with a long-term follow-up, the LIBRA index was significantly predictive for dementia in midlife and late-life participants, but not in the oldest-olds. In people aged 80 and over, only a different model for the calculation of LIBRA scores that included age, sex and education could identify people who subsequently progressed to dementia. Authors observed that their outcomes needed further investigation, since data were available only for 9 or less LIBRA MRF. Our results refer to a dramatically smaller sample and are calculated based on a cross-sectional study design. Given these limitations, they indicate that, once corrected for demographics, LIBRA index seemed not helpful neither in identifying people with increased dementia risk in the whole sample of 308 participants, nor in discriminating between elders with SCD and people with MCI. However, coherently with the cohort study, we identified a relation between LIBRA scores and markers of cognitive decline in the group of younger participants with SCD, while no association between LIBRA and neuropsychological scores was appreciable in the group of older subjects with MCI.

In another cross-sectional study, conducted by Pons and colleagues on a clinical health-seeking sample from community-based adults, aged 45–75 years, that underwent a complete clinical and neuropsychological assessment [23], LIBRA scores showed small to moderate negative correlations with several different cognitive domains. Even in this case, correlations were weaker in older (aged 60 years and over) than in younger participants. Moreover, their MCIp obtained significantly higher LIBRA scores compared SCDp. The authors concluded that LIBRA could be a useful tool to predict the risk of cognitive decline in middle-aged and seniors, even considering the study intrinsic limitations, including the cross-sectional design, the lack of data about food consumption, and the possible selection bias induced by having included health-seeking subjects and not participants from the general population.

Indeed, the LIBRA index was originally developed to estimate an individual’s risk for developing dementia in cognitively preserved adults, not in people with MCI. Our results indicate that, contrarily to those of Pons and colleagues [23], the LIBRA index might not be useful in people with initial objective cognitive impairment. Since our methods and that of Pons and colleagues [23] are largely overlapping, divergent results may be due, at least in part, to sample differences: ours is a community-based sample and it is, on average, older than that recruited by Pons and colleagues [23]. As previously reported, the predictive value of LIBRA appears to decrease as the subject’s age increases. Moreover, since our MCIp are significantly older than SCDp, our results are, again, consistent with the observation that LIBRA index may have a higher predictive power in younger people, while its efficacy in identifying people at risk of dementia decreases with increasing age. In any case, our data seem to agree with the observation that LIBRA scores, once corrected for demographics, are not associated with a greater cognitive decline in MCI patients. Therefore, in people with MCI and in older individuals, other factors, such as age or education, may have a predominant influence compared to MRF.

Conversely, in our study, LIBRA seemed to correctly identify those people who obtained lower neuropsychological test scores among younger participants without objective cognitive decline. Indeed, while no significant association between LIBRA scores and cognitive performances was appreciable in the MCI subgroup when controlling for age and educational levels, a correlation between LIBRA scores and the scores of neuropsychological tests, in particular those measuring global cognitive performances, visuo-spatial attention, and verbal fluency, was detectable in SCDp.

According to some authors, neuropsychological tests show poor accuracy and sensitivity in detecting cognitive difficulties in subjects with SCD and in discriminating individuals that are going to develop dementia from those who are not [50]. Nonetheless, there is some evidence showing that neuropsychological performances and lifestyles may have a role in predicting future cognitive decline in subjects with SCD [51]. SCDp performance in semantic verbal fluency tasks are worse, compared to normal controls, which could be considered an early neuropsychological marker of subtle cognitive impairment [52 –54] and a greater decline in global cognition and language abilities in people with SCD is associated with a more random pattern in brain network organization [55]. Functional connectivity changes have been also observed in SCDp in networks related to visual and attentional tasks [56, 57] and differences exist between the performance of normal controls, SCD, and MCI in attentional tasks, particularly in tests of divided attention [58].

In our opinion, it is noteworthy that the same cognitive areas are those influenced by LIBRA MRF in our subsample of SCDp. It is possible that MRF may play an important role in individuals who are not cognitively impaired, while, in people with MCI, n-MRF might have a major impact on cognitive decline. This may indicate that MRF weighs more on the onset of cognitive disorders than on their progression. When degeneration process is already at an advanced stage, the importance of MRF may be lower compared than that of other factors.

In our study, LIBRA scores were consistently associated with psycho-behavioral markers both in the whole sample and in subsamples of SCDp and MCIp. These behavioral symptoms, especially depression, are well known neuropsychiatric markers of progression to dementia in older people and MCI patients [59]. While data regarding apathy, possibly due to methodological differences among studies, are not univocal [60, 61], several studies indicate that this neuropsychiatric disturbance plays an important role in the progression to dementia [62, 63] However, apathy is both a common symptom of late-life depression and a condition that may present independently from depression in aged individuals with MCI [62]. Apathy is associated with alterations in the frontal-prefrontal circuits [64] and it mediates cognitive difficulties in geriatric patients with depression and cognitive decline, in particular those related to executive functions [63, 65].

In our study, a direct correlation between LIBRA and depression was observed in the whole sample and in the analyses that were carried out separately for SCDp, while the severity of apathy was invariably associated with the entity of the risk determined by the MRF in all samples.

Unfortunately, due to the cross-sectional nature of our study, it is not possible to draw any conclusions on the directionality of the relationship between LIBRA index and cognitive or neuropsychiatric markers. However, GR-2013-02356043 is still in progress, therefore, in the future year we will be able to verify if the LIBRA index may mediate the progression of cognitive, functional, and psycho-behavioral decline and the occurrence of dementia in seniors who have been included in our preventive lifestyle intervention programs and in control subjects, who are not carrying out any further activity to those commonly practiced in their daily life, in order to counteract their age-related cognitive decline.

Taken together, these results may drive future interventions aimed to slow down progression of cognitive decline and possibly, to improve the mental health of older adults. Several randomized controlled trials showed that prevention programs focused on MRF can induce an improvement in cognitive scores [66, 67], and it was suggested that the LIBRA may be a useful index to address methods or effectiveness of lifestyle multidomain interventions. However, a post-hoc analysis of data from the pre-DIVA trial [68, 69]—a multicenter trial aimed to verifying the effectiveness of a 6-year multidomain vascular care intervention to prevent dementia—LIBRA scores were not useful in identifying high-risk individuals who would have benefited most from the intervention. Moreover, despite lifestyle and general health conditions are demonstrated risk factors for depression [70], a recent study showed that lifestyle factors (i.e., physical exercise, sleep disturbance and sleep duration, BMI, and smoking) do not predict the course of depression in older adults [71]. Even this limited evidence should be replicated in studies enrolling seniors at risk of dementia before drawing any conclusion, we should take in account that interventions targeting MRF might neither have a different effect in subjects that are highly exposed to MRF, compared to those with low exposure, nor be useful in reducing depressive symptoms, after depression onset. Moreover, the presence of depression could counteract or weaken the effect of any lifestyle intervention targeting MRF and aimed at the reduction of dementia risk in older persons. Therefore, future MRF-based trials should not underestimate depressive and apathetic symptoms or they should include behavioral interventions aimed at promoting mental health [72].

Strengths and limitations

Beyond its many limitations, the strengths of this study are having included all the LIBRA risk and protective factors, while previous studies have always missed one or more MRFs, in particular Mediterranean diet. In this way, we were the first to obtain a complete picture of the importance of LIBRA MRF on the psycho-behavioral and cognitive state of older people. Moreover, our research enrolled a community-based sample, which is more representative of the general population than a clinical sample. In any case, our participants were recruited among seniors who were willing to be included in a preventive program aimed at reducing dementia risk, while we have no data regarding seniors who were not interested in participating in the trial. Therefore, we should not neglect the existence of a potential selection bias. Moreover, our sample, as well as those of Pons et al. [23], included only subjects with subjective or ascertained cognitive decline, and did not consider cognitively healthy seniors or elders who were not fully aware of their difficulties. Hence, our data should still be confirmed by large population studies. Other limitations of our study can be ascribed to the small sample size and to the cross-sectional design. Therefore, no conclusion can be drawn regarding the causality of the relations between LIBRA and outcomes from the cognitive, functional, and behavioral evaluation. Just as it is possible that MRF lead to worse cognitive and psycho-behavioral status, it is equally possible for cognitive impairment, depression, or apathy to be determinant for a poorer lifestyle. Future longitudinal studies enrolling large samples are therefore needed to confirm the predictive value of the LIBRA index in people at risk for cognitive decline.

Conclusions

Our study demonstrates that LIBRA is significantly associated with depression and apathy in seniors, and with markers of possible future cognitive decline like global cognitive and verbal fluency scores in seniors with SCD. These data support the utility of the LIBRA for the early identification of older individuals at risk for dementia, that may represent proper candidate for preventive personalized lifestyle programs, allowing to intervene when the neurodegenerative processes have not progressed to the point of frustrating any attempt to delay or postpone the onset of dementia.

Footnotes

ACKNOWLEDGMENTS

We would like to address our heartfelt thanks to all the participants, colleagues, and the support staff who contributed to our research. In particular, we want to thank: Dr. Beatrice Pavoni, Dr. Francesca De Masi, Dr. Claudia Natale, Dr. Beatrice Filiputti, and Dr. Matilde Luchi, who participated in the data collection during their internship in Psychology in Santa Lucia Foundation IRCCS; Presidents and participants of Senior Social Centers (SSC) of the Municipalities of Rome for their warm welcome and willingness to take part to the study; Dr. Sergio Longo^†, Dr. Bianca Maria Palleschi, Dr. Maria Elena Raschi, and the other volunteers from the Non-Profit Association “La Casa del Sole” and Dr. Alfonso Rossi and all the staff of the Association for Social Promotion “VitAttiva”, for their collaboration in mediating the participation of SSC Presidents and seniors.

The authors assert that all procedures contributing to this work comply with the ethical standards of the Santa Lucia Foundation IRCCS Ethical Research Committee (Rome, Italy).

Study GR-2013-02356043 was co-financed by the Italian Ministry of Health. Funding sources were not involved in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication.

Authors’ disclosures available online ().

The supplementary material is available in the electronic version of this article: .

References

Prince

, Comas-Herrera

, Knapp

, Guerchet

, Karagiannidou

(2016) World Alzheimer Report 2016: Improving healthcare for people living with dementia. Coverage, Quality and costs now and in the future. Alzheimer’s Disease International, London (UK). https://www.alz.co.uk/research/WorldAlzheimerReport2016.pdf

Stephan

BCM

, Birdi

, Tang

EYH

, Cosco

, Donini

, Licher

, Ikram

, Siervo

, Robinson

(2018) Secular trends in dementia prevalence and incidence worldwide: A systematic review. J Alzheimers Dis 66, 653–680.

Larson

, Yaffe

, Langa

(2013) New insights into the dementia epidemic. N Engl J Med 369, 2275–2277.

Roehr

, Pabst

, Luck

, Riedel-Heller

(2018) Is dementia incidence declining in high-income countries? A systematic review and meta-analysis. Clin Epidemiol 10, 1233–1247.

Hendrie

, Smith-Gamble

, Lane

, Purnell

, Clark

, Gao

(2018) The Association of early life factors and declining incidence rates of dementia in an elderly population of African Americans. J Gerontol B Psychol Sci Soc Sci 73, S82–S90.

Barnes

, Yaffe

(2011) The projected effect of risk factor reduction on Alzheimer’s disease prevalence. Lancet Neurol 10, 819–828.

Norton

, Matthews

, Barnes

, Yaffe

, Brayne

(2014) Potential for primary prevention of Alzheimer’s disease: An analysis of population-based data. Lancet Neurol 13, 788–794.

Petersen

(2004) Mild cognitive impairment as a diagnostic entity. J Intern Med 256, 183–194.

Canevelli

, Grande

, Lacorte

, Quarchioni

, Cesari

, Mariani

, Bruno

, Vanacore

(2016) Spontaneous reversion of mild cognitive impairment to normal cognition: A systematic review of literature and meta-analysis. J Am Med Dir Assoc 17, 943–948.

10.

Petersen

, Smith

, Waring

, Ivnik

, Tangalos

, Kokmen

(1999) Mild cognitive impairment: Clinical characterization and outcome. Arch Neurol 56, 303–08.

11.

Jessen

, Amariglio

, Van Boxtel

, Breteler

, Ceccaldi

, Chételat

, Dubois

, Dufouil

, Ellis

, Van Der Flier

, Glodzik

, Van Harten

, De Leon

, McHugh

, Mielke

, Molinuevo

, Mosconi

, Osorio

, Perrotin

, Petersen

, Rabin

, Rami

, Reisberg

, Rentz

, Sachdev

, De La Sayette

, Saykin

, Scheltens

, Shulman

, Slavin

, Sperling

, Stewart

, Uspenskaya

, Vellas

, Visser

, Wagner

(2014) A conceptual framework for research on subjective cognitive decline in preclinical Alzheimer’s disease. Alzheimers Dement 10, 844–852.

12.

Reisberg

, Shulman

, Torossian

, Leng

, Zhu

(2010) Outcome over seven years of healthy adults with and without subjective cognitive impairment. Alzheimers Dement 6, 11–24.

13.

Studart Neto

, Nitrini

(2016) Subjective cognitive decline: The first clinical manifestation of Alzheimer’s disease? Dement Neuropsychol 10, 170–177.

14.

Valech

, Tort-Merino

, Coll-Padrós

, Olives

, León

, Rami

, Molinuevo

(2017) Executive and language subjective cognitive decline complaints discriminate preclinical Alzheimer’s disease from normal aging. J Alzheimers Dis 61, 689–703.

15.

Tan

C-C

, Yu

J-T

, Wang

H-F

, Tan

M-S

, Meng

X-F

, Wang

, Jiang

, Zhu

X-C

, Tan

(2014) Efficacy and safety of donepezil, galantamine, rivastigmine, and memantine for the treatment of Alzheimer’s disease: A systematic review and meta-analysis. J Alzheimers Dis 41, 615–631.

16.

Barak

, Aizenberg

(2010) Is dementia preventable? Focus on Alzheimer’s disease. Exp Rev Neurother 10, 1689–1698.

17.

Kivipelto

, Ngandu

, Laatikainen

, Winblad

, Soininen

, Tuomilehto

(2006) Risk score for the prediction of dementia risk in 20 years among middle aged people: A longitudinal, population-based study. Lancet Neurol 5, 735–741.

18.

Anstey

, Cherbuin

, Herath

(2013) Development of a new method for assessing global risk of Alzheimer’s disease for use in population health approaches to prevention. Prev Sci 14, 411–421.

19.

Anstey

, Cherbuin

, Herath

, Qiu

, Kuller

, Lopez

, Wilson

, Fratiglioni

(2014) A self-report risk index to predict occurrence of dementia in three independent cohorts of older adults: The ANU-ADRI. PLoS One 9, e86141.

20.

Schiepers

OJG

, Köhler

, Deckers

, Irving

, O’Donnell

, van den Akker

, Verhey

FRJ

, Vos

SJB

, de Vugt

, van Boxtel

MPJ

(2018) Lifestyle for Brain Health (LIBRA): A new model for dementia prevention. Int J Geriatr Psychiatry 33, 167–175.

21.

Deckers

, Van Boxtel

MPJ

, Schiepers

OJG

, De Vugt

, Sánchez

JLM

, Anstey

, Brayne

, Dartigues

, Engedal

, Kivipelto

, Ritchie

, Starr

, Yaffe

, Irving

, Verhey

FRJ

, Köhler

(2015) Target risk factors for dementia prevention: A systematic review and Delphi consensus study on the evidence from observational studies. Int J Geriatr Psychiatry 30, 234–246.

22.

Vos

SJB

, Van Boxtel

MPJ

, Schiepers

OJG

, Deckers

, De Vugt

, Carrière

, Dartigues

, Peres

, Artero

, Ritchie

, Galluzzo

, Scafato

, Frisoni

, Huisman

, Comijs

, Sacuiu

, Skoog

, Irving

, O’Donnell

, Verhey

FRJ

, Visser

, Köhler

(2017) Modifiable risk factors for prevention of dementia in midlife, late life and the oldest-old: Validation of the LIBRA Index. J Alzheimers Dis 58, 537–547.

23.

Pons

, Lamonica

, Mowszowski

, Köhler

, Deckers

, Naismith

(2018) Utility of the LIBRA Index in relation to cognitive functioning in a clinical health seeking sample. J Alzheimers Dis 62, 373–384.

24.

Magni

, Binetti

, Bianchetti

, Rozzini

, Trabucchi

(1996) Mini-mental state examination: A normative study in Italian elderly population. Eur J Neurol 3, 198–202.

25.

Pfeffer

, Kurosaki

, Harrah

, Chance

, Filos

(1982) Measurement of functional activities in older adults in the community. J Gerontol 37, 323–329.

26.

Lawton

, Brody

(1969) Assessment of older people: Self-maintaining and instrumental activities of daily living. Gerontologist 9, 179–186.

27.

Monaco

, Costa

, Caltagirone

, Carlesimo

(2013) Forward and backward span for verbal and visuo-spatial data: Standardization and normative data from an Italian adult population. Neurol Sci 34, 749–754.

28.

Novelli

, Papagno

, Capitani

, Laiacona

, Cappa

, Vallar

(1986) Tre test clinici di memoria verbale a lungo termine. Taratura su soggetti normali. Arch Psicol Neurol Psichiatr 47, 278–296.

29.

Gainotti

(2001) A double dissociation between accuracy and time of execution on attentional tasks in Alzheimer’s disease and multi-infarct dementia. Brain 124, 731–738.

30.

Marra

, Gainotti

, Scaricamazza

, Piccininni

, Ferraccioli

, Quaranta

(2013) The multiple features target cancellation (MFTC): An attentional visual conjunction search test. Normative values for the Italian population. Neurol Sci 34, 173–180.

31.

Caffarra

, Vezzadini

, Dieci

, Zonato

, Venneri

(2002) Rey-Osterrieth complex figure: Normative values in an Italian population sample. Neurol Sci 22, 443–447.

32.

Appollonio

, Leone

, Isella

, Piamarta

, Consoli

, Villa

, Forapani

, Russo

, Nichelli

(2005) The frontal assessment battery (FAB): Normative values in an Italian population sample. Neurol Sci 26, 108–116.

33.

Capasso

, Miceli

(2001) Esame neuropsicologico per l’afasia ENPA. Metodologie riabilitative in logopedia. 4, Springer Science & Business Media.

34.

Novelli

, Papagno

, Capitani

, Laiacona

, Vallar

, Cappa

(1986) Three clinical tests to research and rate the lexical performance of normal subjects. Arch Psicol Neurol Psichiatr 47, 477–506.

35.

Juva

, Mäkelä

, Erkinjuntti

, Sulkava

, Yukoski

, Valvanne

, Tilvis

(1997) Functional assessment scales in detecting dementia. Age Ageing 26, 393–400.

36.

Mejia

, Gutierrez

, Villa

, Ostrosky-Solis

(2005) Cognition, functional status, education, and the diagnosis of dementia and mild cognitive impairment in Spanish-speaking elderly. Appl Neuropsychol 11, 194–201.

37.

Biernat

, Stupnicki

, Lebiedziński

, Janczewska

(2008) Assessment of physical activity by applying IPAQ questionnaire. Phys Educ Sport 52, 46–52.

38.

Martínez-González

MÁ

, Corella

, Salas-Salvadó

, Ros

, Covas

, Fiol

, Wärnberg

, Arós

, Ruíz-Gutiérrez

, Lamuela-Raventós

, Lapetra

, Muñoz

MÁ

, Martínez

, Sáez

, Serra-Majem

, Pintó

, Mitjavila

, Tur

, del Puy Portillo

, Estruch

(2012) Cohort profile: Design and methods of the PREDIMED study. Int J Epidemiol 41, 377–385.

39.

Schröder

, Fitó

, Estruch

, Martínez-González

, Corella

, Salas-Salvadó

, Lamuela-Raventós

, Ros

, Salaverría

, Fiol

, Lapetra

, Vinyoles

, Gómez-Gracia

, Lahoz

, Serra-Majem

, Pintó

, Ruiz-Gutierrez

, Covas

(2011) A short screener is valid for assessing Mediterranean diet adherence among older Spanish men and women. J Nutr 141, 1140–1145.

40.

Hoyl

, Alessi

, Harker

, Josephson

, Pietruszka

, Koelfgen

, Mervis

, Fitten

, Rubenstein

(1999) Development and testing of a five-item version of the geriatric depression scale. J Am Geriatr Soc 47, 873–878.

41.

Marin

, Biedrzycki

, Firinciogullari

(1991) Reliability and validity of the apathy evaluation scale. Psychiatry Res 38, 143–162.

42.

Oveisgharan

, Hachinski

(2010) Hypertension, executive dysfunction, and progression to dementia: The Canadian study of health and aging. Arch Neurol 67, 187–192.

43.

Garrow

, Webster

(1985) Quetelet’s index (W/H2) as a measure of fatness. Int J Obes 9, 147–153.

44.

Rehm

, Room

, Taylor

(2008) Method for moderation: Measuring lifetime risk of alcohol-attributable mortality as a basis for drinking guidelines. Int J Methods Psychiatr Res 17, 141–151.

45.

De Amicis

, Leone

, Foppiani

, Osio

, Lewandowski

, Giustizieri

, Cornelio

, Fusari Imperatori

, Cappa

, Battezzati

, Bertoli

(2018) Mediterranean diet and cognitive status in free-living elderly: A cross-sectional study in Northern Italy. J Am Coll Nutr 37, 494–500.

46.

Hernández-Galiot

, Goñi

(2017) Adherence to the Mediterranean diet pattern, cognitive status and depressive symptoms in an elderly non-institutionalized population. Nutr Hosp 34, 338–344.

47.

Leone

, Battezzati

, De Amicis

, De Carlo

, Bertoli

(2017) Trends of adherence to the Mediterranean dietary pattern in northern Italy from 2010 to 2016. Nutrients 9, E734.

48.

Bertoli

, Leone

, Vignati

, Bedogni

, Martínez-González

MÁ

, Bes-Rastrollo

, Spadafranca

, Vanzulli

, Battezzati

(2015) Adherence to the Mediterranean diet is inversely associated with visceral abdominal tissue in Caucasian subjects. Clin Nutr 34, 1266–1272.

49.

Soldati

, Bertoli

, Terranegra

, Brasacchio

, Mingione

, Dogliotti

, Raspini

, Leone

, Frau

, Vignati

, Spadafranca

, Vezzoli

, Cusi

, Battezzati

(2014) Relevance of Mediterranean diet and glucose metabolism for nephrolithiasis in obese subjects. J Transl Med 12, 34.

50.

Jessen

(2014) Subjective and objective cognitive decline at the pre-dementia stage of Alzheimer’s disease. Eur Arch Psychiatry Clin Neurosci 264, 3–7.

51.

Bessi

, Mazzeo

, Padiglioni

, Piccini

, Nacmias

, Sorbi

, Bracco

(2018) From subjective cognitive decline to Alzheimer’s disease: The predictive role of neuropsychological assessment, personality traits, and cognitive reserve. A 7-year follow-up study. J Alzheimers Dis 63, 1523–1535.

52.

Nikolai

, Bezdicek

, Markova

, Stepankova

, Michalec

, Kopecek

, Dokoupilova

, Hort

, Vyhnalek

(2018) Semantic verbal fluency impairment is detectable in patients with subjective cognitive decline. Appl Neuropsychol 25, 448–457.

53.

Clark

, Gatz

, Zheng

, McCleary

, MacK

, Chen

(2009) Longitudinal verbal fluency in normal aging, preclinical, and prevalent Alzheimer’s disease. Am J Alzheimers Dis Other Demen 24, 461–468.

54.

Murphy

, Rich

, Troyer

(2006) Verbal fluency patterns in amnestic mild cognitive impairment are characteristic of Alzheimer’s type dementia. J Int Neuropsychol Soc 12, 570–574.

55.

Verfaillie

SCJ

, Slot

RER

, Dicks

, Prins

, Overbeek

, Teunissen

, Scheltens

, Barkhof

, van der Flier

, Tijms

(2018) A more randomly organized grey matter network is associated with deteriorating language and global cognition in individuals with subjective cognitive decline. Hum Brain Mapp 39, 3143–3151.

56.

Wang

, Qiao

, Chen

, Sui

, Dong

H-M

, Wang

Y-S

, Li

H-J

, Lu

, Zuo

X-N

, Han

(2019) Functional connectivity changes across the spectrum of subjective cognitive decline, amnestic mild cognitive impairment and Alzheimer’s disease. Front Neuroinform 13, 1–11.

57.

Rodda

, Dannhauser

, Cutinha

, Shergill

, Walker

(2011) Subjective cognitive impairment: Functional MRI during a divided attention task. Eur Psychiatry 26, 457–462.

58.

, Lo

, Huang

, Deng

, Hsu

(2018) Visual attention performances and related cerebral microstructural integrity among subjects with subjective cognitive decline and mild cognitive impairment. Front Aging Neurosci 10, 1–13.

59.

Kim

, Kang

, Kim

, Ryu

, Kim

, Seo

, Na

, Han

(2013) Neuropsychiatric predictors of conversion to dementia both in patients with amnestic mild cognitive impairment and those with subcortical vascular MCI. Clin Neurol Neurosurg 115, 1264–1270.

60.

A. Köhler

, F. Magalhaes

, M.M.P. Oliveira

, S. Alves

, Knochel

, Oertel-Knöchel

, Pantel

, F. Carvalho

(2016) Neuropsychiatric disturbances in mild cognitive impairment (MCI): A systematic review of population-based studies. Curr Alzheimer Res 13, 1066–1082.

61.

Brodaty

, Heffernan

, Draper

, Reppermund

, Kochan

, Slavin

, Trollor

, Sachdev

(2012) Neuropsychiatric symptoms in older people with and without cognitive impairment. J Alzheimers Dis 31, 411–420.

62.

Palmer

, Di Iulio

, Varsi

, Gianni

, Sancesario

, Caltagirone

, Spalletta

(2010) Neuropsychiatric predictors of progression from amnestic – Mild cognitive impairment to alzheimer’s disease: The role of depression and apathy. J Alzheimers Dis 20, 175–183.

63.

Funes

, Lavretsky

, Ercoli

, St. Cyr

, Siddarth

(2018) Apathy mediates cognitive difficulties in geriatric depression. Am J Geriatr Psychiatry 26, 100–106.

64.

Moretti

, Signori

(2016) Neural correlates for apathy: Frontal-prefrontal and parietal cortical- subcortical circuits. Front Aging Neurosci 8, 289.

65.

Perri

, Turchetta

, Caruso

, Fadda

, Caltagirone

, Carlesimo

(2018) Neuropsychological correlates of cognitive, emotional-affective and auto-activation apathy in Alzheimer’s disease. Neuropsychologia 118, 12–21.

66.

Ngandu

, Lehtisalo

, Solomon

, Levälahti

, Ahtiluoto

, Antikainen

, Bäckman

, Hänninen

, Jula

, Laatikainen

, Lindström

, Mangialasche

, Paajanen

, Pajala

, Peltonen

, Rauramaa

, Stigsdotter-Neely

, Strandberg

, Tuomilehto

, Soininen

, Kivipelto

(2015) A 2 year multidomain intervention of diet, exercise, cognitive training, and vascular risk monitoring versus control to prevent cognitive decline in at-risk elderly people (FINGER): A randomised controlled trial. Lancet 385, 2255–2263.

67.

Andrieu

, Coley

, Lovestone

, Aisen

, Vellas

(2015) Prevention of sporadic Alzheimer’s disease: Lessons learned from clinical trials and future directions. Lancet Neurol 14, 926–944.

68.

Van Charante

EPM

, Richard

, Eurelings

, van Dalen

, Ligthart

, van Bussel

, Hoevenaar-Blom

, Vermeulen

, van Gool

(2016) Effectiveness of a 6-year multidomain vascular care intervention to prevent dementia (preDIVA): A cluster-randomised controlled trial. Lancet 388, 797–805.

69.

Van Middelaar

, Hoevenaar-Blom

, Van Gool

, Moll Van Charante

, Van Dalen

, Deckers

, Köhler

, Richard

(2018) Modifiable dementia risk score to study heterogeneity in treatment effect of a dementia prevention trial: A post hoc analysis in the preDIVA trial using the LIBRA index. Alzheimers Res Ther 10, 62.

70.

Tanaka

, Sasazawa

, Suzuki

, Nakazawa

, Koyama

(2011) Health status and lifestyle factors as predictors of depression in middle-aged and elderly Japanese adults: A seven-year follow-up of the Komo-Ise cohort study. BMC Psychiatry 11, 20.

71.

Bruin

, Comijs

, Kok

, Van der Mast

, Van den Berg

(2018) Lifestyle factors and the course of depression in older adults: A NESDO study. Int J Geriatr Psychiatry 33, 1000–1008.

72.

Kane

, Butler

, Fink

, Brasure

, Davila

, Desai

, Jutkowitz

, McCreedy

, Nelson

, McCarten

, Calvert

, Ratner

, Hemmy

, Barclay

(2017) Interventions to Prevent Age-Related Cognitive Decline, Mild Cognitive Impairment, and Clinical Alzheimer’s-Type Dementia. Agency for Healthcare Research and Quality (US); 2017 Mar. Report No.: 17-EHC008-EF. Rockville (MD).