Dementia prevention, intervention, and care

TL;DR: The Lancet Commission on Dementia Prevention, Intervention, and Care met to consolidate the huge strides that have been made and the emerging knowledge as to what the authors should do to prevent and manage dementia.

About: This article is published in The Lancet.The article was published on 2017-12-16 and is currently open access. It has received 3826 citations till now. The article focuses on the topics: Dementia & Long-term care.

Summary

INTRODUCTION

• Natural ecosystems, overall biodiversity, human health and natural economies are affected at global, regional and local scales by an ever-increasing number of destructive alien species (Vitousek et al., 1997; Pimentel et al., 2001).
• This relationship assumes that a species’ current distribution provides useful information regarding the species’ environmental requirements (Pearson et al., 2007).
• When occurrence records are geographically biased, the underlying environmental gradients in which a species can persist will most likely also not be fully sampled, which could result in environmental bias (Raes & ter Steege, 2007; Hortal et al., 2008).

MATERIALS AND METHODS

• Both countries have good sources of distribution records that are readily available through electronic databases.
• For the Australian species invading South Africa, native range occurrence records were obtained from the Australian Virtual Herbarium public access database (AVH; http://www.ersa.edu.au/avh/, accessed 15 February 2007).
• In both cases, all native range records were assumed to be representative of the entire native range of the species concerned.
• Only one record per grid cell was used when several occurrence records were present in a grid cell (Hernandez et al., 2006).

Environmental predictors

• The authors selected the 19 bioclimatic variables available from the WorldClim database .
• Bioclimatic variables derived from monthly temperature and precipitation data are commonly used in biogeographical modelling (De Meyer et al., 2008; Loiselle et al., 2008).
• These variables represent annual trends (e.g. annual mean temperature), seasonality (e.g. annual range in precipitation) and extreme or limiting environmental factors (e.g. precipitation of the wettest month) (Hijmans et al., 2005).
• All environmental predictors were resampled to 15¢ grids using ArcGIS 9.2.

Simulating biased sampling

• The authors started by defining the full range of a species as all of the native range distribution records available for that species.
• This represents an extreme case where no records are available for the species from a specific part of the range.
• For scenario B, an initial sample of 50% of the records from either the east or south of the range was used and then a further 40% of the remaining records (from the west or north, respectively) was randomly sampled and added to the initial sample (B10).
• Datasets with the same number in the code (e.g. A10, B10 and R10) contained equal numbers of occurrence records.

Ecological niche modelling

• Maxent is a relatively new statistical modelling technique that has been applied to model the potential distribution of species and to estimate niche occupation (Phillips et al., 2006; Peterson et al., 2008; Phillips & Dudı́k, 2008).
• Even when absence data are available, they are usually unreliable, as the species may have been recorded as absent merely because insufficient time had elapsed to allow for invasion (Wilson et al., 2007; Peterson et al., 2008).
• This is a continuous variable ranging from 0 to 1, where high values indicate higher suitability for a species in a particular grid cell (Phillips et al., 2006).
• All models were calibrated with samples of records taken from the native range (the calibration set) (Fig. 2).
• Algorithm parameters were set to a maximum number of 500 iterations, a regularization multiplier of 1, auto features and a convergence threshold of 0.00001.

Ecological niches

• Comparisons of marginality values between the native and introduced range indicated that there was no significant difference in the position of the niche occupied between these two ranges for any of the species (Table 2).
• A comparison of the tolerance values between these two ranges indicated that the environmental variation between the native and introduced range differed significantly for 12 of the 19 species (Table 2).
• Values indicated in bold were significantly different (P < 0.05) between the different ranges.
• Differences in environmental bias across treatments Environmental bias, expressed as the difference in marginality between treatment and control datasets for all species , differed significantly across treatments (H = 26.13, P < 0.05).
• When measured as the difference in tolerance , environmental bias showed no significant differences across treatments (H = 10.18, P = 0.069).

Model evaluation

• Model performance is frequently assessed using the area under the curve (AUC) of receiver-operating characteristic (ROC) plots (Fielding & Bell, 1997; Lobo et al., 2008; Peterson et al., 2008).
• Recently, the use of AUC statistics for model evaluation has been criticized (Lobo et al., 2008).
• This method plots the true positive fraction as a function of the proportion of the overall area predicted present.
• This evaluation method eliminates the reliance on commission error, where areas might be classified as unsuitable based on uncertain absences, i.e. pseudo-absences (Peterson et al., 2008).
• TSS values range from )1 to +1, where +1 indicates a perfect fit and values of 0 or less indicate a performance no better than random (Allouche et al., 2006).

Analysis

• In order to describe the relationship between geographical and environmental bias the authors had to quantify these biases.
• The marginality and tolerance value for a dataset comprising biased records is likely to be smaller than for a dataset comprising randomly selected records, as a biased dataset will tend to sample less environmental variation.
• The authors show how the bias created in the occurrence records relates to the geographical and environmental distribution of the species .
• To assess whether the amount of environmental bias differed on average across treatments the authors compared the difference in marginality or difference in tolerance values for all the species between a control dataset and the relevant bias dataset (e.g. R10 and A10).
• Nonparametric comparisons were carried out with Kruskal–Wallis tests, and then the significant differences were identified with multiple comparisons carried out with the npmc function in R (R Development Core Team, 2004).

DISCUSSION

• Relationship between geographical bias and environmental bias Values of the test statistic obtained when each model performance measure was compared between a control and biased datasets (P > 0.05).
• In other words, the centres of the niches occupied in the native and introduced ranges were in the same part of the environmental space, but the amount of environmental variation differed significantly between the native and introduced ranges.
• When making predictions of introduced range for alien invasive species it may be necessary to calibrate models with native range records.

Figures

Figure 13. RCTs of effect of psychosocial interventions vs controls for agitation in dementia; standardised effect size and 95% confidence intervals, where calculable, for agitation immediately and in the longer term

Figure 11. Approaches to assessing and managing agitation in dementia (see text for further details)

Figure 4. Life course model of contribution of modifiable risk factors to dementia

Table 2. Potentially modifiable risk factors for progression to dementia from MCI; their relative risk, prevalence and population attributable fraction

Figure 1. The growth in numbers of people with dementia in high-income and low- and middleincome countries

Figure 15. Effect of psychological treatment vs treatment as usual; outcome is depression

Figure 17. Approaches to assessing mental capacity

Figure 2. Brain autopsy results from cognitively normal individuals expressed as Summary Neuropathology Score (potentially range 0 to 9) ranked from lowest to highest.

Table 1. Risk factors for dementia; relative risk for dementia, prevalence, communality and Population Attributable Fraction (PAF) for dementia

Figure 8. Effect of memantine at optimum dose on cognition, measured by the severe impairment battery, in moderate to severe Alzheimer’s disease; mean change in score from baseline at 6 months or later

Figure 9. Effect of Cognitive Stimulation Therapy vs usual care on cognition, measured by the MMSE

Figure 12. Guidance for using pleasant activity as a therapeutic intervention in dementia

Figure 6. Effect of cholinesterase inhibitors at optimum dose on cognition, measured by MMSE, in

Figure 10. Approaches to assessing and managing psychosis in dementia (see text for further details)

Figure 16. Effect of 10 days to 10 weeks of bright light treatment on total sleep duration

Figure 3. Forest plot of effect of hearing loss, measured by pure tone audiometry, on incident dementia rates 9-17 years later in cognitively healthy people

Figure 18. Model of palliative care in dementia; from European Association of Palliative Care White Paper

Figure 5. Potential brain mechanisms for preventative strategies in dementia

Frequently asked questions

Q1. What contributions have the authors mentioned in the paper "The lancet international commission on dementia prevention and care" ?

In this paper, the authors present a plan for the future of dementia care in the context of the International Commission on Dementia Prevention and Care ( ICPC ). 

Q2. What could be done to prevent dementia?

More childhood education, exercise, maintaining social engagement, reducing or stopping smoking, management of hearing loss, depression, diabetes, hypertension and obesity could all contribute to prevention or delay of dementia. 

Q3. What is the purpose of this report?

The authors have itemised interventions which can transform the lives of people with dementia and their families, maximising cognition, decreasing distressing associated symptoms, reducing crises and improving quality of life. 

Q4. How does the author propose a new life-course model of risk?

Incorporating potentially reversible risk factors from different phases of the life-span and not just old age, the authors are able to propose a novel life-course model of risk from which population attributable fractions have been derived to demonstrate the possible impact on future incidence of successful elimination of the most potent factors. 

Q5. How does the doctor deal with dementia?

Dementia affects the individual living with it, who gradually loses abilities, as well as their relatives and other supporters, who have to cope with seeing a family member or friend become ill and decline, while responding to their needs, such as increasing dependency and changes in behaviour. 

Q6. What is the role of dementia in the care of older people?

A third of older people now die with dementia and all professionals working in endof-life care need to make this a central part of their planning and communication. 

Q7. How many people are living with dementia?

There were around 47 million people living with dementia globally in 2015, and this is projected toincrease to 66 million by 2030 and 115 million by 2050. 

Q8. What is the role of dementia prevention and care?

Effective dementia prevention and care could transform the future for society and vastly improve living and dying for individuals with dementia and their families. 

Q9. What can be done to prevent dementia?

If these are implemented, people with dementia will have their cognition optimised and be less likely to be agitated, depressed or have troublesome psychotic symptoms and family carers will have reduced levels of anxiety and depression. 

Q10. What is the main point of the report?

Nonetheless delaying dementia for some years for even a small percentage of people would be an enormous achievement and enable many more people to reach the end-of-life without developing dementia. 

Q11. What is the way to prevent dementia?

Many people present to services with Mild Cognitive Impairment (MCI) a risk state for dementia, which occurs in up to a fifth of people aged over 65 and this provides an opportunity for more targeted interventions. 

Q12. How many people with dementia are preventable?

The authors have brought together all this evidence and have calculated that around one third of dementia may theoretically be preventable.