Utility of the LIBRA Index in Relation to Cognitive Functioning in a Clinical Health Seeking Sample

Abstract

Background:

Dementia prevalence is expected to increase substantially over the next few decades. Since there is currently no cure for dementia available, there is an urgent need for the early identification of individuals at high risk for dementia, so that primary and secondary prevention strategies can be implemented. Recently, the LIfestyle for BRAin health (LIBRA) index was developed as a new dementia risk algorithm. It specifically focuses on modifiable risk and protective factors that can be targeted in midlife.

Objective:

The objective of this study was to evaluate the LIBRA index in relation to markers of cognitive functioning in a clinical, health-seeking sample of community-based older adults.

Methods:

484 participants (mean age 62.7 years) were recruited from the Healthy Brain Ageing Clinic at the Brain and Mind Centre, Sydney. Participants underwent comprehensive clinical and neuropsychological assessment and completed a self-report survey pack. Participants were rated via consensus as having either subjective cognitive complaints (SCC) or meeting criteria for mild cognitive impairment (MCI). The LIBRA score was calculated based on 11 available risk and protective factors.

Results:

65.4% of the sample met criteria for MCI. People with MCI showed a significantly higher LIBRA score compared to people with SCC. Furthermore, multiple cognitive domains, in particular executive functioning, were associated with a higher LIBRA score, with stronger correlations in people with MCI.

Conclusion:

The LIBRA index might be a useful tool to determine lifestyle-attributable risk of cognitive decline in an older health-seeking population, including people with MCI.

Keywords

Dementia mild cognitive impairment modifiable risk factors prevention

INTRODUCTION

In 2015, it was estimated that 46.8 million people worldwide were affected by dementia, a number predicted to almost double every 20 years, reaching 131.5 million people by 2050 [1]. The global costs of dementia in 2015 were estimated at US$ 818 billion [2]. Besides this huge economic burden, dementia also greatly impacts on quality of life for individuals living with dementia as well as their family and carers. Dementia is rated number 9 in the top 10 most burdensome conditions among older people worldwide [1], mainly impacted by years lived with disability, rather than years of life lost from premature mortality. Furthermore, when examining dependence (the need for care), dementia and cognitive impairment are both the leading chronic disease contributors, and are also associated with more intensified need for care compared to other chronic conditions [3].

As dementia is currently incurable, there is a high need for effective preventive strategies [4]. In this regard, early identification of individuals at high risk for dementia is crucial, and there is a need for better identification of those highest at-risk along with tailored management of modifiable dementia risk factors. Of significance, in a meta-analytic study, Norton at al. showed that, after adjusting for non-independence between risk factors, 30% of the population attributable risk for Alzheimer’s disease (AD) is related to seven potentially modifiable risk factors, including educational attainment, vascular risk factors, and depression [5]. More recently, Ashby et al. estimated that 48% of dementia cases in Australia could be attributed to the same seven modifiable risk factors [6]. They also showed that a reduction of these modifiable risk factors by 5–20% per decade would reduce the prevalence of future dementia in Australia by between 6.6 and 30.7% in 2050. Their findings also showed that the percentage of cases related to modifiable risk factors was higher than those related to APOE ɛ4 (the main genetic risk factor for AD), again highlighting the importance of targeting modifiable risk factors in dementia prevention. This overall approach emphasizing risk reduction is supported by data showing that there has been a 22% decrease in dementia prevalence in the United Kingdom due to earlier population-level investments such as prevention and treatment of vascular conditions, and improved education and living conditions [4]. These findings suggest that primary prevention will have the largest effect on reduction of future dementia occurrence and disability. Starting early in life or midlife seems most promising for primary prevention, as 1) this is likely early enough to intervene before neurodegeneration and cognitive changes have commenced [7, 8], and 2) since the effects of risk factors change over the course of life, with some risk factors having a bigger magnitude in midlife than in later life (including obesity and hypertension) [9 –11].

Of note, selection of individuals that may benefit most from prevention programs is important as treatment effects might not be equally distributed along the spectrum from low to high initial risk [12]. Early identification of individuals at higher risk for dementia may be achieved by the development and implementation of dementia risk algorithms. Several prediction models to calculate an individual’s dementia risk have been developed. The most extensively researched of these is the Cardiovascular Risk Factors, Aging, and Incidence of Dementia (CAIDE) index, which aims to predict a 20-year dementia risk based on midlife vascular risk factors [13]. However, although the CAIDE index has been validated in external datasets, it is based on a single cohort study, making it less generalizable and less suitable for global identification of individuals at risk for dementia [12]. To overcome this, the Australian National University Alzheimer’s Disease Risk Index (ANU-ADRI) has been developed as a self-report dementia risk assessment tool based on a literature review rather than single-cohort studies [14]. However, the ANU-ADRI, like other dementia risk scores, comprises both modifiable and non-modifiable risk factors such as age or sex, making it less suitable for prevention strategies.

Recently, the LIfestyle for BRAin health (LIBRA) index [15] was developed, which has an emphasis on modifiable risk factors that can be targeted for prevention efforts [16]. The LIBRA index was developed from a systematic literature review on the epidemiological evidence for dementia risk factors, followed by expert Delphi consensus [16]. The resulting LIBRA index identified 12 modifiable risk and protective factors for dementia. The relative risk of each factor was standardized and weighted to a reference value used to calculate a personalized LIBRA global score. In a large multicenter study of a European population-based cohort, Vos et al. reported that the LIBRA index based on eight risk factors and one protective factor could be a useful tool to identify individuals at risk for dementia in midlife (55–69 years) and late-life (70–79 years), but not in the oldest old (>80 years) [17]. Schiepers et al. examined LIBRA with respect to prospective longitudinal epidemiological data and again confirmed the predictive utility of LIBRA for dementia risk [15]. Specifically, an increase in the global LIBRA score by one point was related to a 19% higher risk for developing dementia. Again however, LIBRA appears to be of less value in people aged 75 years and older [17, 18].

To date, however, the LIBRA index has only been evaluated in community-based population studies. No research has been conducted to evaluate how the LIBRA index relates to sensitive markers of cognitive decline in a clinical, health-seeking population, nor in a population with well characterized neuropsychological markers of cognitive decline. Given that LIBRA might be a valuable tool in clinical settings too, it is important to establish the conditions under which LIBRA may show the greatest association with early cognitive change. Therefore, the aims of our study were: 1) to determine whether the LIBRA global score is associated with neuropsychological functioning in a midlife to older age sample attending a specialist early intervention clinic for subjective cognitive complaints; 2) to determine whether the association between the LIBRA global score and neuropsychological functioning is more pronounced in midlife (45–59) compared to older age (60–75); 3) to determine whether the LIBRA global score is associated with cognitive status (subjective complaints versus mild cognitive impairment (MCI)) in a sample seeking assessment for memory concerns; and, 4) to determine whether the association between the LIBRA global score and cognition is more pronounced in people with MCI.

METHODS

Participants

Participants aged between 45 and 75 years were recruited from the Healthy Brain Ageing (HBA) Clinic at the Brain and Mind Centre, University of Sydney, Sydney, Australia. This is a specialist assessment and early intervention clinic for health-seeking people reporting subjective concerns about cognition and/or mood. Exclusion criteria were: limited English proficiency, Mini-Mental State Examination (MMSE) <20, intellectual disability as defined in the Diagnostic and Statistical Manual – Fourth Edition (DSM-IV) [19], history of stroke, traumatic brain injury (with loss of consciousness >30 min), neurological or other medical conditions known to affect cognition, current substance misuse, or major non-affective psychiatric disorder (e.g., psychosis). We excluded participants with missing data on more than three from the total of twelve risk/protective factors to allow calculation of LIBRA scores. Additionally, as LIBRA was developed as an algorithm to calculate an individual’s dementia risk score, participants with an existing diagnosis of dementia were also excluded.

Participation was voluntary and written informed consent was obtained from all participants prior to the clinical assessment. This study received ethical approval from the University of Sydney Human Research Ethics Committee.

Assessments

As described previously [20], all participants underwent a comprehensive clinical assessment as follows:

Medical assessment

Prior to attending the HBA clinic, participants were asked to obtain a referral letter from their GP or specialist outlining their medical history and medication overview. A medical specialist (geriatrician and/or neurologist) performed a semi-structured clinical assessment including cognitive and medical history, current treatments, family history, sleep disturbances, and current and past history of substance use, including smoking history. A neurological examination was undertaken and biomedical assessments included weight, height, waist circumference and blood pressure (measured twice, sitting down). Additionally, the physician administered the MMSE [21] to screen for global cognitive dysfunction, the Instrumental Activities of Daily Living and Physical Self-Maintenance Scale [22] and the Activities of Daily Living Scale [23] to assess functional status, and The Cumulative Illness Rating Scale – Geriatric version [24] as a measure of medical burden. In relation to the latter, 14 organ systems (e.g., heart, vascular, hematopoietic, respiratory, etc.) were rated on a scale of 0 to 4, with 0 reflecting ‘no problem’ and 4 indicating ‘extremely severe/immediate treatment required/end organ failure/severe impairment in function’.

Mood assessment

All participants underwent a mood assessment with a trained research psychologist. A set of structured interview tools were used to assess history and current symptoms of depression and anxiety, including the Mini-International Neuropsychiatric Interview (M.I.N.I), a short-structured interview screening for all major Axis 1 disorders according to DSM-IV diagnostic criteria. The Major Depression module of the M.I.N.I was used to assess lifetime and current symptoms of major depression [25].

Neuropsychological assessment

A clinical neuropsychologist conducted a standardized battery of neuropsychological tests, assessing the following cognitive domains:

Processing speed: This domain was assessed using the Trail Making Test, Part A (TMT-A) [26] and the Delis-Kaplan Executive Functioning System (D-KEFS) Color-Word Interference Test, Color Naming and Word Reading subtests [27].

Working memory: The Digit Span (DS) subtest of the Wechsler Adult Intelligence Scale-Third Edition (WAIS-III) [28] was used to examine participants’ working memory based on the number of digits repeated in both forward and backward sequence.

Verbal learning: The Rey Auditory Verbal Learning Test (RAVLT) [29], a serial list-learning test, and the Logical Memory (LM) subtest of the Wechsler Memory Scale, Third Edition (WMS-III) [28], were used to assess encoding of unstructured and structured verbal information, respectively.

Memory: Percent retention scores from the RAVLT (trial 7/5) and LM tests were used to measure delayed recall.

Language: The Boston Naming Test (BNT) [30] was used as a measure of confrontation naming and the Controlled Oral Word Association Test (COWAT) [31] was used to assess semantic fluency (animals).

Executive functions: The Trail Making Test, Part B (TMT-B) was used to evaluate mental flexibility and D-KEFS Color-Word Interference Test Inhibition and Inhibition/Switching subtests were used to assess response inhibition. Additionally, the COWAT was used to measure phonemic fluency.

Visuospatial functioning: Visuospatial abilities were indexed using the Rey-Osterrieth Complex Figure Test Copy (RCFT-Copy) [32].

Premorbid IQ: Premorbid intellectual functioning was estimated using the Wechsler Test of Adult Reading (WTAR) [33]. The WTAR score and total years of education were used as proxy markers of cognitive reserve.

Normative data were used to calculate z-scores for all tests, and averaged z-scores were calculated for each cognitive domain. For TMT A and B, a raw score that is lower than the mean from the normative data results in a negative z-score. Therefore, the sign of all z-scores for TMT A and B were reversed upon entry into the SPSS database to facilitate statistical analyses.

Self-report questionnaires

Participants were asked to complete several self-report questionnaires prior to their appointment in the clinic, assessing physical activity levels, mood, sleep, and cognitive stimulation. Physical activity was assessed using the self-report Active Australia Survey, which has been widely used as a measure of physical activity [34]. Participants reported frequency and duration of walking, moderate physical activity, and vigorous physical activity over the last week. To evaluate depressive symptoms, participants completed the Geriatric Depression Scale-short form (GDS-15), a 15-item self-report measure that has been demonstrated to be a valid screening instrument for elderly adults [35]. A cut-score of ≥6 is suggestive of depression. Sleep was assessed using the Pittsburgh Sleep Quality Index (PSQI) [36], a commonly used measure of sleep quality which has good reliability and validity in both non-clinical and clinical samples [37], and in which a higher score indicates poorer sleep quality. Participants also completed a self-report measure assessing the degree to which they engaged in cognitively stimulating activities (Cognitively Stimulating Activities, CSA) (e.g., reading the newspaper, playing card games, doing paper-based brain activities) over the past month. In total, 13 different activities were rated on a scale ranging from 1 ‘never’ to 5 ‘every day or about every day.’ The sum of these 13 questions provided the CSA total score.

Diagnostic procedure

Diagnoses of MCI and dementia were determined via consensus of three raters including a neurologist or geriatrician and two clinical neuropsychologists. History, clinical presentation, functional status, neuropsychological performance, and neuroimaging findings (when available) were all taken into account in the diagnostic process. Participants were rated as having MCI if they had a decline of at least 1.5 SD on one or more standardized measures of neuropsychological functioning relative to their estimated baseline level of performance (age adjusted standard score), in addition to self-reported cognitive decline [38], but in the absence of significant functional impairment. Established diagnostic criteria were used in the differential diagnosis of dementia. Participants who did not meet criteria for MCI or dementia were rated as participants with subjective cognitive complaints (SCC).

Definition of LIBRA risk and protective factors

Data was available for 11 of the 12 LIBRA factors. In case of missing items, the LIBRA score was calculated based on the remaining items if at least 9 risk and protective factors were present.

Risk factors

Depression

For this study, participants scored positive for the risk factor ‘depression’ if they reported either a lifetime history of depression on the M.I.N.I. or a GDS score ≥6.

Physical inactivity

Based on recommendations from the National Physical Activity Guidelines for Australians [39], sufficient activity is defined as at least 150 minutes of activity and at least five sessions of activity over one week. Participants who did not meet these criteria, were rated as physically inactive.

Smoking

Research shows that only current smoking is associated with increased risk of AD and possibly other dementias compared to a smoking history [40]; therefore, only current smokers had a positive rating on this risk factor.

Hypercholesterolemia

Participants with a history of hypercholesterolemia based on their referral letter, self-report and/or as revealed in the medical assessment were coded risk factor positive.

(Midlife) hypertension

A history of hypertension was obtained from the referral letter, with more details regarding age of onset, treatment, and current blood pressure gathered during the medical assessment. Given evidence that only midlife hypertension is associated with higher risk of dementia/cognitive decline, participants were rated positive when the onset of hypertension was between the ages of 40 and 60 years, or their repeated systolic blood pressure measurement was ≥140 mmHg and/or diastolic blood pressure ≥90 mmHg when aged 40 to 60 years at the time of assessment.

(Midlife) obesity

During the medical assessment, weight and height were measured and BMI was calculated. Participants with a BMI ≥30 were rated positive for this risk factor. While obesity is only included as a risk factor in the LIBRA index when present in midlife, it was included as risk factor for all participants regardless of age in the present study.

Coronary heart disease

The risk factor ‘Coronary heart disease’ was rated based on the CIRS item regarding heart problems: 0 ‘no problems’, 1 ‘remote myocardial infarction (MI) (>5 years ago)/occasional angina treated with medication’, 2 ‘CCF compensated with medications/daily anti-angina medication/left ventricular hypertrophy/atrial fibrillation/bundle branch block/daily antiarrhythmic drugs’, 3 ‘previous MI within 5 years/abnormal stress test/status post percutaneous coronary angioplasty or coronary artery bypass graft surgery’, 4 ‘marked activity restriction secondary to cardiac status (i.e., unstable angina or intractable congestive heart failure). Any score other than 0 was rated as risk factor positive.

Renal dysfunction

The CIRS was also used to assess renal problems with any score >1 being coded as risk factor positive. The potential scores included: 0 ‘No problem’, 1 ‘kidney stone passage within the past 10 years or asymptomatic kidney stone/pyelonephritis within 5 years’, 2 ‘Serum creatinine >1.5 but <3.0 without diuretic or antihypertensive medication’, 3 ‘Serum creatinine >3.0 or serum creatinine >1.5 in conjunction with diuretic, antihypertensive, or bicarbonate therapy/current pyelonephritis’, 4 ‘requires dialysis/renal carcinoma’.

Diabetes

Information regarding diabetes was collected from the referral letter, self-report and during the medical assessment. Only participants diagnosed with type 2 diabetes were rated as risk factor positive.

Protective factors

Low/moderate alcohol intake

Alcohol intake was assessed during the interview with the medical specialist. Participants were asked about their current alcohol intake in drinks per day or drinks per week. Based on the National Health and Medical Research Council (NHMRC) guidelines, low/moderate alcohol intake was rated as <2 standard drinks per day or <14 standard drinks per week [41]. No alcohol intake was not rated as a protective factor.

High cognitive activity

This item was rated based on the CSA questionnaire score and total years of education, as assessed by the neuropsychologist. Participants were rated ‘highly cognitively active’ when the combined score of the CSA questionnaire and total years of education was within the highest tertile of the total sample.

Mediterranean diet

Insufficient data were available to rate this protective factor.

Statistical analysis

Data were analyzed using IBM SPSS Statistics version 24. Descriptive statistics were used to analyze the demographic characteristics of the study sample and to describe the mean questionnaire/neuropsychological test scores. Independent two tailed t-tests (α <0.05) were used to compare characteristics between diagnostic groups (i.e., different age groups; participants with SSC versus MCI) and partial correlations were used to examine the relationship between LIBRA score and medical or psychological measures corrected for age (note: age was accounted for in neuropsychological measures through the use of normative data, as outlined above). As there were no significant differences in sex and education level, scores were not corrected for these factors. Nine participants had one of their neuropsychological test scores curtailed due to outliers (to maximum -3.0 SD).

RESULTS

As of June 2016, 729 participants consented to undergo a baseline assessment in the HBA clinic. A total 245 participants were excluded (mean age 72.9 years (SD 9.6), 53.1% female) (73 due to insufficient data on the risk and protective factors associated with the LIBRA global score (mean age 65.2 years (SD 9.7), 41.1% female, 43.8% MCI); 66 participants due to a diagnosis of dementia (mean age 71.4 years (SD 9.5), 56.1% female, mean LIBRA score 2.3 (SD 2.5)); and 106 participants because of age >75 (mean age 78.8 years (SD 4.7), 54.7% female, 78.3% MCI, mean LIBRA score 2.7 (SD 2.8)). The remaining 484 participants included in this study had a mean age of 62.7 years (SD 7.12). The majority of participants were female (58.1%) and met criteria for MCI (65.4%). The average LIBRA score was 2.2 (SD 2.8, range –4.2–10.5), of which 46.5% had data on all 11 risk factors, 30.4% on 10 risk factors and 23.1% on 9 risk factors. Cognitively stimulating activity was the main missing factor (84.2%).

Table 1 shows the demographic data in relation to the diagnostic group. The participants with MCI were slightly older, had a higher medical burden, prevalence of depression and average LIBRA score compared to participants with SCC. With regard to cognition, participants with MCI had lower scores on all neuropsychological tests compared with participants to SCC, with the exception of premorbid intellectual functioning (WTAR).

Table 1

Sample descriptives for patients with subjective cognitive complaints (SCC) and mild cognitive impairment (MCI)

	SCC (n = 168)	MCI (n = 316)	t-value	p–value
Age, y	61.8±7.3 (45.0–74.9)	63.2±7.0 (45.0–74.9)	–2.01	0.04
Sex
Female	104 (61.9%)	178 (56.2%)	χ² = 1.56	0.22
Male	64 (38.1%)	139 (43.8%)
Education, y	13.9±3.0 (4.0–21.0)	13.7±3.2 (6.0–22.0)	0.63	0.53
Body Mass Index	27.0±5.4 (18.0–48.0)	27.3±5.5 (17.2–49.9)	–0.60	0.59
MMSE	29.3±1.0 (25–30)	28.6±1.7 (22–30)	5.59	<0.01
GDS	3.6±3.8 (0–15)	5.1±4.1 (0–15)	–3.76	<0.01
CIRS	3.6±3.0 (0–17)	4.9±3.9 (0–26)	–4.07	<0.01
PSQI	6.4±3.9 (0–17)	7.1±3.9 (0–18)	–1.79	0.08
LIBRA score	1.81±2.64 (–4.2–10.5)	2.38±2.86 (–4.2–10.1)	–2.11	0.03
Neuropsychological tests
WTAR predicted IQ	105.6±10.6 (14–120)	104.9±10.0 (70–126)	0.61	0.53
Digit Span (WMS)^*	0.62±1.06 (–2.66–2.33)	0.03±1.33 (–3.00–2.33)	5.31	<0.01
Logical memory I (WMS)^*	0.55±0.95 (–2.33–2.66)	–0.21±1.29 (–3.00–2.00)	7.29	<0.01
Logical memory % retention (WMS)^*	0.78±0.86 (–2.00–2.00)	0.30±1.15 (–3.00–2.00)	5.10	<0.01
COWAT FAS^*	0.41±1.00 (–2.21–3.00)	–0.16±0.96 (–2.48–0.89)	6.08	<0.01
COWAT Animals^*	0.63±0.99 (–2.19–3.00)	0.02±1.17 (–2.85–3.00)	5.98	<0.01
RAVLT trail 1 to 5^*	0.38±0.74 (–2.09–2.17)	–0.37±1.00 (–3.00–1.98)	8.54	<0.01
RAVLT 20-min delay % retention^*	0.20±0.80 (–3.00–1.55)	–0.36±1.24 (–3.00–1.84)	5.56	<0.01
Trails A^*	0.43±0.78 (–1.95–1.99)	–0.04±1.04 (–3.00–1.80)	5.57	<0.01
Trials B^*	0.40±0.67 (–1.86–1.87)	–0.27±1.12 (–3.00–2.05)	8.17	<0.01
DKEFS Stroop 4^*	0.59±0.63 (–2.00–1.66)	0.08±1.10 (–3.00–2.00)	6.19	<0.01
BNT^*	0.67±1.05 (–2.66–2.33)	0.05±1.33 (–3.00–2.33)	5.47	<0.01

Mean±SD (range) SCC, subjective cognitive complaints; MCI, mild cognitive impairment; MMSE, Mini-Mental State Examination, GDS, Geriatric Depression Scale; CIRS, Cumulative Illness Rating Scale; PSQI, Pittsburgh Sleep Quality Index; LIBRA score, LIfestyle for BRAin health score; WTAR, Wechsler Test of Adult Reading; WMS, Wechsler Memory Scale; COWAT, Controlled Oral Word Association Test; RAVLT, Rey Auditory Verbal Learning Test; DKEFS, Delis Kaplan Executive Functioning System; BNT, Boston Naming Test; χ², Pearson Chi-Square value ^*z-score.

Table 2 demonstrates the correlations between the LIBRA score and the different cognitive domains in relation to age, in order to distinguish between midlife (aged 45–59) and older age (aged 60–75). As expected, medium to large positive correlations were found between the LIBRA score and depression and medical burden, with stronger correlations in midlife compared to older age. Cognitive reserve was negatively correlated with the LIBRA score in both age groups, indicating that higher premorbid intellect is associated with a lower dementia risk. The strongest correlation between the different cognitive domains and the LIBRA score was found for executive functioning, with a higher negative correlation in midlife compared to older age. Small significant negative correlations were found between the LIBRA score and processing speed and learning in midlife, and processing speed, working memory and language in older age. It is also important to note that all correlations, including the non-significant correlations, were directionally similar.

Table 2

Correlations between LIBRA score and different cognitive domains per age group

	Entire sample n = 484	Midlife (Age 45–59) n = 156	Older age (Age 60–75) n = 328
Age, y	0.01	–0.07	–0.02
Depression (GDS-15)^†	0.39^**	0.44^**	0.37^**
Medical burden (CIRS)^†	0.48^**	0.58^**	0.42^**
MMSE^†	–0.10^*	–0.18^*	–0.07
Cognitive reserve markers
Education, y	–0.36^**	–0.25^**	–0.40^**
WTAR predicted IQ	–0.26^**	–0.32^**	–0.23^**
Cognitive domains
Processing speed	–0.21^**	–0.24^**	–0.19^**
Working memory	–0.18^**	–0.12	–0.21^**
Learning	–0.14^**	–0.23^**	–0.11
Memory	–0.07	–0.04	–0.08
Language	–0.12^**	–0.14	–0.11^*
Visuospatial functioning	–0.03	–0.06	–0.01
Executive functioning	–0.28^**	–0.43^**	–0.21^**

GDS, Geriatric Depression Scale; CIRS, Cumulative Illness Rating Scale; MMSE, Mini-Mental State Examination; WTAR, Wechsler Test of Adult Reading. ^†corrected for age, ^*p < 0.05, ^**p < 0.01.

The correlations between the LIBRA score and the different cognitive domains in relation to diagnostic groups are shown in Table 3. Small to medium positive correlations were found between the LIBRA score and depression and medical burden, with stronger correlations in the MCI group, demonstrating that people with MCI and depression or a high medical burden have a higher risk of developing dementia. The correlations between the LIBRA score and the cognitive reserve markers were similar in both diagnostic groups. In people with MCI, processing speed, working memory and executive functioning showed small negative correlations with the LIBRA score.

Table 3

Correlations between LIBRA score and cognitive domains according to cognitive impairment group (ages 45–75)

	Entire sample N = 484	Participants with SCC N = 168	Participants with MCI N = 316
Age, y	0.01	0.08	–0.04
Depression (GDS-15)^†	0.39^**	0.21^**	0.40^**
Medical burden (CIRS)^†	0.48 ^**	0.42 ^**	0.49 ^**
MMSE^†	–0.10*	–0.13	–0.09
Cognitive reserve markers
Education, y	–0.36^**	–0.36^**	–0.35^**
WTAR predicted IQ	–0.26^**	–0.22^**	–0.27^**
Cognitive domains
Processing speed	–0.21^**	–0.01	–0.24^**
Working memory	–0.18^**	–0.16*	–0.16^**
Learning	–0.14^**	–0.23^**	–0.08
Memory	–0.07	–0.20*	–0.01
Language	–0.12^**	–0.11	–0.10
Visuospatial functioning	–0.03	–0.22*	–0.02
Executive functioning	–0.28^**	–0.16	–0.29^**

SCC: subjective cognitive complaints, MCI: mild cognitive impairment, GDS: Geriatric Depression Scale, CIRS: Cumulative Illness Rating Scale, MMSE: Mini-Mental State Examination, WTAR: Wechsler Test of Adult Reading. ^†corrected for age, ^*p < 0.05, ^**p < 0.01.

To clarify the relationship between the LIBRA score and cognition in people with MCI compared to participants with SCC in the different age groups, correlations between the different domains and the LIBRA score were determined per diagnostic group for people in midlife and in older-aged people. Table 4 demonstrates that depression and medical burden were positively correlated with the LIBRA score in both diagnostic groups. In midlife, the significant correlations found in Table 2, were only seen in participants with MCI, not in participants with SCC, whereas in the older aged group, significant correlations were found in both people with MCI and SCC.

Table 4

Correlations between LIBRA score and cognitive domains in midlife and older age, per diagnostic group

	Midlife (age 45–59)		Older age (age 60–75)
	Participants with SCC n = 59	Participants with MCI n = 97	Participants with SCC n = 109	Participants with MCI n = 219
Age	–0.23	–0.04	0.13	–0.04
Depression (GDS-15)^†	0.52^**	0.37^**	0.21^*	0.41^**
Medical burden (CIRS)^†	0.42^**	0.63^**	0.42^**	0.42^**
MMSE^†	–0.25	–0.13	–0.08	–0.08
Cognitive reserve markers
Education, years	–0.12	–0.29^**	–0.46^**	–0.38^**
WTAR predicted IQ	–0.19	–0.36^**	–0.23^*	–0.24^**
Cognitive domains
Processing speed	–0.13	–0.24^*	0.05	–0.24^**
Working memory	–0.02	–0.13	–0.22^*	–0.19^**
Learning	–0.16	–0.19	–0.25^*	–0.04
Memory	–0.10	–0.02	–0.23^*	–0.02
Language	–0.03	–0.13	–0.15	–0.08
Visuospatial functioning	–0.02	–0.08	–0.26^*	0.04
Executive functioning	–0.16	–0.49^**	–0.16	–0.21^**

SCC, subjective cognitive complaints; MCI, mild cognitive impairment; GDS, Geriatric Depression Scale; CIRS, Cumulative Illness Rating Scale; MMSE, Mini-Mental State Examination; WTAR, Wechsler Test of Adult Reading ^†corrected for age, ^*p < 0.05, ^**p < 0.01.

A sensitivity analysis was performed to evaluate the impact of the risk factor ‘depression’ on the described correlations. The risk factor depression was removed from the LIBRA global score, and all correlations were corrected for depression instead. The resulting analyses did not show significant changes to the original results, demonstrating that the correlations were not confounded by the risk factor depression.

DISCUSSION

This is the first study to evaluate the LIBRA index in a clinical, health-seeking sample. Our findings show that people with MCI have a significantly higher LIBRA global score compared to people with SCC. Furthermore, the LIBRA global score shows consistently small to moderate negative correlations with several different cognitive domains in people with MCI. This suggests that the LIBRA index might be a useful tool to determine risk of cognitive decline in an older health-seeking population, including people with MCI.

Interestingly, the negative correlations shown between the LIBRA global score and most of the different cognitive domains, were all stronger in midlife (45–59) compared to late life (60–75). This confirms the findings by Vos et al., who showed the index is mainly useful to predict dementia for people in midlife, whereas in later life an alternative risk-prediction model based on both modifiable and non-modifiable (age, sex, and education) risk factors is likely to be superior [17].

The LIBRA index was originally developed to estimate an individual’s risk for developing dementia in cognitively intact adults; however, our results indicate the index might also be useful in people with MCI. In both diagnostic groups (i.e., people with SCC and people with MCI), consistent negative correlations were found between the LIBRA score and the cognitive domains. MCI has been used to describe the transition between normal ageing and dementia, and is considered to be a preclinical stage of dementia [42], although around 40% of people with MCI can remain stable or even revert to previous levels of functioning [43]. Nonetheless, annual conversion rates are much higher than that seen in a normal population, ranging from 10–15% [38] in clinical samples to 4–10% in community samples [44]. As people with MCI have a higher risk of developing dementia compared to people with normal cognition [45], and MCI is a period during which (cognitive) interventions are thought to have a beneficial effect on decreasing the likelihood of dementia progression, it is important to identify individuals within this group with a high risk of conversion. In a recent review, Kirova et al. state that executive dysfunction and working memory impairments appear to be a sign of progression to dementia [46]. Our results align with this statement, as they show executive functioning to be associated with a greater risk of dementia, with stronger correlations in people with MCI. Kirova et al. also suggest that therapeutic interventions aimed to stop or slow the progression to dementia should focus on tasks that emphasize executive function, as it has been shown that neuroplasticity and cognitive restructuring in the brain as a result of these interventions can decrease the risk of developing dementia [47]. For example, it has been demonstrated that Problem-Solving Therapy (PST), a behavioral intervention targeting executive deficits, has beneficial effects on depression and cognitive impairments in older adults with depression [48, 49]. Cognitive training, particularly programs involving strategy-based approaches, have also been shown to have moderate sized effects when targeting executive functions [50]. Other interventions proven to be effective in decreasing the risk of developing dementia include cognitive training programs [51], physical exercise [52] and cardiovascular health [53, 54].

The LIBRA index has previously been shown to be predictive of dementia and cognitive decline in a single longitudinal study as well as in a multicenter cohort study. The average LIBRA score in both studies were slightly different compared to our mean LIBRA score (2.2). Schiepers et al. showed a slightly lower LIBRA score of 1.6, a difference possibly explained by the fact that our most common lacking LIBRA factor was cognitive activity, which is a protective factor and therefore would lower the total LIBRA score. The LIBRA score of 2.9 shown by Vos et al. was higher compared to our LIBRA score, probably due to their sample being considerably older on average than our population, and potentially having a higher medical burden, for example. Our study was not able to determine predictive ability, as this was a cross-sectional study. Nevertheless, our results do align with prior work as we were able to show a higher LIBRA score in participants with MCI as well as significant negative correlation between the LIBRA score and several cognitive domains, indicating a higher LIBRA score is associated with lower cognitive functioning. When comparing our results to previously described indices, our study is one of the first to look at a clinical, health-seeking population. In a recent study, the ANU-ADRI showed strong associations with progression from cognitively normal to MCI; however, the predictive ability was limited, possibly due to the relatively young age of the cohort and consequently the small number of participants with MCI [14]. An important difference between the LIBRA index and the previously described indices is that it does not include non-modifiable risk factors, including age, which has the largest weighting of risk factors in the CAIDE index and the ANU-ADRI. This might make the LIBRA index more useful as a preventive tool, whereas the others might be more useful as a predictive tool.

New studies are required now to investigate the potential of the LIBRA index to respond to early intervention and prevention efforts. O’Donnell et al. are currently conducting the In-MINDD study, a multi-center trial exploring the feasibility and acceptability of an online LIBRA global risk profile and an online support environment to help individuals assess and possibly reduce their risk of developing dementia [55]. Previous research has shown that online interventions (e-health) are effective in the treatment of various medical and mental health conditions [56, 57]. Although very limited research has been conducted evaluating the utility of e-health technologies for the prevention of MCI and/or dementia, our recent study shows that the vast majority (92.8%) of older adults with cognitive concerns use a computer routinely and in particular, they report using the Internet to access health-related websites [58]. Furthermore, this work shows that such individuals are interested in online interventions targeting cognition and memory as well as other risk factors for cognitive decline. These results highlight the potential of online prevention models such as the LIBRA index.

It is important to acknowledge the limitations of this study when interpreting the results. First, this study was conducted in a clinical sample, whom are (by virtue of the intended purpose of the clinic) enriched for cognitive decline. However, while these results may not generalize to the broader population, it is noted that the European studies have already examined the utility of LIBRA in population studies. Also, the clinic setting is where the index might be most useful, as these people seek assessment for cognitive concerns. Second, as this study was a cross-sectional study, is was not possible to assess temporality of associations, nor was it possible to rule out reverse causations (i.e., cognitive impairment could lead to poorer lifestyle). However, our findings were in line with previous prospective studies on LIBRA in the community. Third, the Mediterranean diet ‘protective’ factor was not available in our cohort, but it is worth noting that studies and instruments used to assess adherence to this diet are often of poor quality and specification [59]. In addition, several risk factors were not collected on all participants. In accordance with previous studies on LIBRA, at least 9 risk factors were required to calculate an individual’s LIBRA score, however it is possible this lack of a complete (12-item) LIBRA score may have influenced the results. Information on all risk factors is now implemented in the HBA clinic for future research. Fourth, our participants had a relatively high level of education and were largely residing in the metropolitan Sydney area and, therefore, may not be representative of the general community in Australia. Fifth, no significant correlations were found in the smallest group (participants with SCC in midlife), indicating the LIBRA score is a better predictor in people with MCI in this age group. However, it is important to note alternative explanations for this observation, as it could also be explained by a small sample size, or the fact that in this young group the neuropsychological tests are insensitive to very subtle preclinical decline, making it harder to detect any significant correlations. Finally, as this was an exploratory study in nature, we chose not to correct for multiple comparisons in the statistical analyses. We acknowledge that this may have increased the likelihood of Type I error; however, this is unlikely given the number of significant correlations found.

In summary, this study is the first to show the LIBRA index is associated with cognitive decline in a clinical sample, and our results indicate the LIBRA index might be a useful tool to identify people at risk for developing dementia in a health-seeking population, including those who meet criteria for MCI. Future studies are required to calculate the predictive ability of the LIBRA index in this population.

Footnotes

ACKNOWLEDGMENTS

We thank the participants who contribute to our research at the Healthy Brain Ageing clinic. Funding sources had no involvement 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. The authors assert that all procedures contributing to this work comply with the ethical standards of the University of Sydney Human Ethical Research Committee (Sydney, Australia).

Authors’ disclosures available online ().

References

Prince

, Prina

, Guerchet

(2015) World Alzheimer Report 2015: The Global Impact of Dementia. Alzheimer’s Disease International, London.

Brown

, Hansnata

, Anh La

(2017) Economic cost of dementia in Australia 2016-2056.

Prince

, Prina

, Guerchet

(2013) World Alzheimer Report 2013: Journey of Caring. Alzheimer’s Disease International, London.

, Fratiglioni

, Matthews

, Lobo

, Breteler

, Skoog

, Brayne

(2016) Dementia in western Europe: Epidemiological evidence and implications for policy making. Lancet Neurol 15, 116–124.

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.

Ashby-Mitchell

, Burns

, Shaw

, Anstey

(2017) Proportion of dementia in Australia explained by common modifiable risk factors. Alzheimers Res Ther 9, 11.

Morris

(2005) Early-stage and preclinical Alzheimer disease. Alzheimer Dis Assoc Disord 19, 163–165.

Sperling

, Aisen

, Beckett

, Bennett

, Craft

, Fagan

, Iwatsubo

, Jack

Jr , Kaye

, Montine

, Park

, Reiman

, Rowe

, Siemers

, Stern

, Yaffe

, Carrillo

, Thies

, Morrison-Bogorad

, Wagster

, Phelps

(2011) Toward defining the preclinical stages of Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 280–292.

Anstey

, Cherbuin

, Budge

, Young

(2011) Body mass index in midlife and late-life as a risk factor for dementia: A meta-analysis of prospective studies. Obes Rev 12, e426–437.

10.

Skoog

, Lernfelt

, Landahl

, Palmertz

, Andreasson

, Nilsson

, Persson

, Oden

, Svanborg

(1996) 15-year longitudinal study of blood pressure and dementia. Lancet 347, 1141–1145.

11.

Whitmer

, Sidney

, Selby

, Johnston

, Yaffe

(2005) Midlife cardiovascular risk factors and risk of dementia in late life. Neurology 64, 277–281.

12.

Solomon

, Kivipelto

, Soininen

(2013) Prevention of Alzheimer’s disease: Moving backward through the lifespan. J Alzheimers Dis 33, S465–S469.

13.

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.

14.

Andrews

, Eramudugolla

, Velez

, Cherbuin

, Easteal

, Anstey

(2017) Validating the role of the Australian National University Alzheimer’s Disease Risk Index (ANU-ADRI) and a genetic risk score in progression to cognitive impairment in a population-based cohort of older adults followed for 12 years. Alzheimers Res Therapy 9, 16.

15.

Schiepers

, Kohler

, Deckers

, Irving

, O’Donnell

, van den Akker

, Verhey

, Vos

, de Vugt

, van Boxtel

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

16.

Deckers

, van Boxtel

, Schiepers

, de Vugt

, Munoz Sanchez

, Anstey

, Brayne

, Dartigues

, Engedal

, Kivipelto

, Ritchie

, Starr

, Yaffe

, Irving

, Verhey

, Kohler

(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.

17.

Vos

SJB

, van Boxtel

MPJ

, Schiepers

OJG

, Deckers

, de Vugt

, Carriere

, Dartigues

, Peres

, Artero

, Ritchie

, Galluzzo

, Scafato

, Frisoni

, Huisman

, Comijs

, Sacuiu

, Skoog

, Irving

, O’Donnell

, Verhey

FRJ

, Visser

, Kohler

(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.

18.

Deckers

, Köhler

, van Boxtel

, Verhey

, Brayne

, Fleming

(2017) Lack of associations between modifiable risk factors and dementia in the very old: Findings from the Cambridge City over-75s cohort study. Aging Ment Health. doi: 10.1080/13607863.2017.1280767

19.

American Psychiatric Association (1994), Diagnostic and statistical manual of mental disorders, Washington, DC.

20.

Duffy

, Lagopoulos

, Hickie

, Diamond

, Graeber

, Lewis

, Naismith

(2014) Glutathione relates to neuropsychological functioning in mild cognitive impairment. Alzheimers Dement 10, 67–75.

21.

Folstein

, Folstein

, McHugh

(1975) “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

22.

Lawton MPB

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

23.

Sheikh

, Smith

, Meade

, Goldenberg

, Brennan

, Kinsella

(1979) Repeatability and validity of a modified activities of daily living (ADL) index in studies of chronic disability. Int Rehab Med 1, 51–58.

24.

Miller

, Paradis

, Houck

, Mazumdar

, Stack

, Rifai

, Mulsant

, Reynolds

(1992) Rating chronic medical illness burden in geropsychiatric practice and research: Application of the Cumulative Illness Rating Scale. Psychiatry Res 41, 237–248.

25.

Sheehan

, Lecrubier

, Sheehan

, Amorim

, Janavs

, Weiller

, Hergueta

, Baker

, Dunbar

(1998) The Mini-International Neuropsychiatric Interview (M.I.N.I.): The development and validation of a structured diagnostic psychiatric interview for DSM-IV and ICD-10. J Clin Psychiatry 59(Suppl 20), 22–33; quiz 34-57.

26.

Tombaugh

(2004) Trail Making Test A and B: Normative data stratified by age and education. Arch Clin Neuropsychol 19, 203–214.

27.

Delis

, Kaplan

, Kramer

(2001) Delis-Kaplan executive function system (D-KEFS), Psychological Corporation.

28.

Wechsler

(1997) Wechsler memory scale (WMS-III), Psychological corporation San Antonio, TX.

29.

Strauss

, Sherman

, Spreen

(2006), A compendium of neuropsychological tests: Administration, norms, and commentary, American Chemical Society.

30.

Ivnik

, Malec

, Smith

, Tangalos

, Petersen

(1996) Neuropsychological tests’ norms above age 55: COWAT, BNT, MAE token, WRAT-R reading, AMNART, STROOP, TMT, and JLO. Clin Neuropsychol 10, 262–278.

31.

Ruff

, Light

, Parker

, Levin

(1996) Benton controlled oral word association test: Reliability and updated norms. Arch Clin Neuropsychol 11, 329–338.

32.

Osterrieth

(1993) The complex figure copy test. Clin Neuropsychol 7, 3–21.

33.

Wechsler

(2001) Wechsler Test of Adult Reading: WTAR, Psychological Corporation.

34.

Australian Institute of Health and Welfare (2003) The active Australia survey: A guide and manual for implementation, analysis and reporting.

35.

Yesavage

, Sheikh

(1986) 9/Geriatric Depression Scale (GDS). Clin Gerontol 5, 165–173.

36.

Buysse

, Reynolds

, Monk

, Berman

, Kupfer

(1989) The Pittsburgh sleep quality index: A new instrument for psychiatric practice and research. Psychiatry Res 28, 193–213.

37.

Mollayeva

, Thurairajah

, Burton

, Mollayeva

, Shapiro

, Colantonio

(2016) The Pittsburgh sleep quality index as a screening tool for sleep dysfunction in clinical and non-clinical samples: A systematic review and meta-analysis. Sleep Med Rev 25, 52–73.

38.

Petersen

, Smith

, Waring

, Ivnik

, Tangalos

, Kokmen

(1999) Mild cognitive impairment: Clinical characterization and outcome. Arch Neurol 56, 303–308.

39.

Sims

, Hill

, Hunt

, Haralambous

, Brown

, Engel

, Huang

, Kerse

, Ory

(2006) National physical activity recommendations for older Australians: Discussion document. Australian Government Department of Health and Ageing, Canberra.

40.

Peters

, Poulter

, Warner

, Beckett

, Burch

, Bulpitt

(2008) Smoking, dementia and cognitive decline in the elderly, a systematic review. BMC Geriatrics 8, 36.

41.

National Health Medical Research Council (2009) Australian guidelines to reduce health risks from drinking alcohol. Commonwealth of Australia Canberra.

42.

Janoutová

, Serý

, Hosák

, Janout

(2015) Is mild cognitive impairment a precursor of Alzheimer’s disease? Cent Eur J Public Health 23, 365–367.

43.

Larrieu

, Letenneur

, Orgogozo

, Fabrigoule

, Amieva

, Le Carret

, Barberger-Gateau

, Dartigues

(2002) Incidence and outcome of mild cognitive impairment in a population-based prospective cohort. Neurology 59, 1594–1599.

44.

Farias

, Mungas

, Reed

, Harvey

, DeCarli

(2009) Progression of mild cognitive impairment to dementia in clinic-vs community-based cohorts. Arch Neurol 66, 1151–1157.

45.

Petersen

, Doody

, Kurz

, Mohs

, Morris

, Rabins

, Ritchie

, Rossor

, Thal

, Winblad

(2001) Current concepts in mild cognitive impairment. Arch Neurol 58, 1985–1992.

46.

Kirova

, Bays

, Lagalwar

(2015) Working memory and executive function decline across normal aging, mild cognitive impairment, and Alzheimer’s disease. BioMed Res Int 2015, 748212.

47.

Huckans

, Hutson

, Twamley

, Jak

, Kaye

, Storzbach

(2013) Efficacy of cognitive rehabilitation therapies for mild cognitive impairment (MCI) in older adults: Working toward a theoretical model and evidence-based interventions. Neuropsychol Rev 23, 63.

48.

Mackin

, Nelson

, Delucchi

, Raue

, Byers

, Barnes

, Satre

, Yaffe

, Alexopoulos

, Arean

(2014) Cognitive outcomes after psychotherapeutic interventions for major depression in older adults with executive dysfunction. Am J Geriatr Psychiatry 22, 1496–1503.

49.

Areán

, Raue

, Mackin

, Kanellopoulos

, McCulloch

, Alexopoulos

(2010) Problem-solving therapy and supportive therapy in older adults with major depression and executive dysfunction. Am J Psychiatry 167, 1391–1398.

50.

Mowszowski

, Lampit

, Walton

, Naismith

(2016) Strategy-based cognitive training for improving executive functions in older adults: A systematic review. Neuropsychol Rev 26, 252–270.

51.

Diamond

, Mowszowski

, Cockayne

, Norrie

, Paradise

, Hermens

, Lewis

, Hickie

, Naismith

(2015) Randomized controlled trial of a healthy brain ageing cognitive training program: Effects on memory, mood, and sleep. J Alzheimers Dis 44, 1181–1191.

52.

Nagamatsu

, Handy

, Hsu

, Voss

, Liu-Ambrose

(2012) Resistance training promotes cognitive and functional brain plasticity in seniors with probable mild cognitive impairment. Arch Intern Med 172, 666–668.

53.

Leshner

, Landis

, Stroud

, Downey

(2017), Preventing Cognitive Decline and Dementia: A Way Forward, Washington, DC.

54.

Ngandu

, Lehtisalo

, Solomon

, Levälahti

, Ahtiluoto

, Antikainen

, Bäckman

, Hänninen

, Jula

, Laatikainen

(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.

55.

O’Donnell

, Browne

, Pierce

, McConnachie

, Deckers

, van Boxtel

, Manera

, Köhler

, Redmond

, Verhey

(2015) Reducing dementia risk by targeting modifiable risk factors in mid-life: Study protocol for the Innovative Midlife Intervention for Dementia Deterrence (In-MINDD) randomised controlled feasibility trial. Pilot Feasibility Stud 1, 40.

56.

Andersson

, Cuijpers

(2009) Internet-based and other computerized psychological treatments for adult depression: A meta-analysis. Cogn Behav Ther 38, 196–205.

57.

Glozier

, Christensen

, Naismith

, Cockayne

, Donkin

, Neal

, Mackinnon

, Hickie

(2013) Internet-delivered cognitive behavioural therapy for adults with mild to moderate depression and high cardiovascular disease risks: A randomised attention-controlled trial. PLoS One 8, e59139.

58.

LaMonica

, English

, Hickie

, Ip

, Ireland

, West

, Shaw

, Mowszowski

, Glozier

, Duffy

, Gibson

, Naismith

(2017) Internet and eHealth practices and preferences of Australian older adults with subjective memory complaints, mild cognitive impairment, or dementia. J Med Internet Res 19, e358.

59.

Radd-Vagenas

, Kouri-Blazos

, Singh

, Flood

(2017) Evolution of Mediterranean diets and cuisine: Concepts and definitions. Asia Pac J Clin Nutr 26, 749–763.