Dementia: New Projections and Time for an Updated Response

This issue of the journal contains an article by Jorm, Dear and Burgess that projects the future number of dementia cases in Australia between 2000 and 2050. The projections include prevalence and incidence rates if no intervention occurs, and if preventative methods could be used to delay the onset of dementia by various time periods. Furthermore, the article describes the likely ratio between healthy workers and dementia cases across the next few decades. Jorm et al. carefully present the numbers, the method is well supported, and the conclusions conservative. Why then is this article the focus of the current editorial? The answer is twofold. First, there is a need to highlight the significant message Jorm et al.'s work sends. Second, it is important to initiate discussion about the validity and completeness of the projections (negative and positive) and what action is required.

Importance of the message

We must first acknowledge that although Jorm et al.'s article provides updated projections, the message is not new. The importance of the article is in emphasizing the recurring nature of the message. Almost 20 years ago, Jorm and Korten published a paper outlining projections for dementia cases in Australia [1]. This paper informed us that between 1984 and 2021, dementia caseswould rise 136–139%, that is, at three times the rate of population growth. The accuracy of Henderson and Jorm's (1986) [2] 20-year projection for the prevalence of dementia adds confidence in accepting the figures given in the new projections. They estimated that there were between 97 800 and 115 000 people with dementia in Australia in 1986, and that there would be between 173 300 and 194 200 in 2006. We had been warned. Jorm et al.'s updated projections paint an equally dismal picture, predicting a 241% rise in dementia cases between 2000 and 2050. Worse still, it appears this country will have difficulty in financing, staffing and caring for these people, given that the projected healthy worker to people with dementia ratio is likely to decrease from 75:1 to 26:1 by 2050. The prediction may, in fact, be conservative: diagnoses of dementia are being made earlier; thresholds of diagnosis are being reconsidered; and it may be possible that the lifespan of people with dementia can be prolonged by treatment. The importance is further underscored by the economic impact of dementia. Access Economics recently calculated that the annual direct and indirect costs of dementia in Australia exceeded $A6bn (Access Economics, 2003 report, www.alzheimers.org.au) [3].

Is the message complete?

The scholarly analysis by Jorm et al. does not take some other factors into consideration. First, there are societal changes that will impact upon the care of dementia suffers in ways that do not easily permit an economic analysis. Two key changes that will lead to a reduced number, and perhaps quality, of family carers for people with dementia are the reduction in family size and the increasing intergenerational age gap as women delay their entry into motherhood. The social as well as economic burden on the family will increase. Furthermore, a greater proportion of families will need to assist in the care of their older relatives, with an impact on family dynamics as well as the workforce. Families may not experience the intergenerational roles usually assigned to healthy families. Adults may leave or reduce workforce hours to care for a parent with dementia or to fill gaps in day care for their own children in the absence of healthy grandparents who would normally have assisted with this.

Jorm et al. rely on the most widely accepted incidence and prevalence rates of dementia, based on meta-analyses of scores of studies, that is, they use the best data available. Without faulting their methodology, we wish to draw attention to the limitations in the definition of dementia that most epidemiological studies overlook. Two recent studies of the reliability and validity of diagnostic criteria of dementia highlight this point. The first study examined 1879 patients 65 years and over from the Canadian Study of Health and Aging [4] and found that the proportion of patients receiving a diagnosis of dementia varied from 3.1% using ICD-10 criteria to 29.1% using DSM-III criteria, a 10-fold difference! An investigation of the reliability and validity of the diagnosis of vascular dementia [5] reported that the concordance between different pairs of criteria varied from 18% to 48%, and much of the variation was because of different definitions of dementia even when these were being used by the same group of investigators. The seeming solidity of the figures may well hide a soft underbelly.

The problem is accentuated by the large penumbra of predementia syndromes, often categorized as mild cognitive impairment (MCI) and its variants. The prevalence rates of this syndrome in the elderly have been estimated as between 3% and 54% depending upon the methodology and criteria used [6]. It is certainly many times commoner than dementia. Even though the burden of having mild impairment is much less than that of dementia, the high prevalence makes this not an inconsiderable problem in the community, which may impact on age of retirement and the productivity of the older citizens. Society must eventually come to terms with the impact of cognitive disorders over and above that of dementia.

Hope

The ultimate hope of a dementia researcher is to be able to prevent cognitive deterioration altogether. Jorm et al. are more modest in their projections, assuming the ability, with future interventions, to delay the onset of dementia by 5 years which would reduce prevalence by 44%, or even as little as 6months which would reduce it by 6%. The idea of being able to prevent dementia, or delay its onset, is not altogether fanciful. It is not so long ago that vascular dementia was hailed as a ‘preventable senility’ [7]. Targeting stroke prevention by controlling risk factors such as hypertension, diabetes, high cholesterol, smoking and atrial fibrillation have been suggested as the most potent strategies for the prevention of vascular dementia [8]. The optimal treatment of hypertension reduces the incidence of stroke as well as small vessel disease. Additionally, hypertension is accepted as a risk factor for Alzheimer's disease, and its control potentially has a dual benefit. The Syst-Eur trial provided preliminary evidence for the effectiveness of the strategy of dual targeting of the common dementias with the control of hypertension [9]. Many other risk and protective factors have been identified for Alzheimer's disease [10]. Attention to high cholesterol, homocysteine and low folic acid are referred to by Jorm et al. Activities that increase brain reserve such as physical exercise and intellectual pursuits such as reading, solving puzzles and other stimulating pastimes have a protective role and have received much attention [11]. Drugs such as anti-oxidants, antiinflammatories and statins have been suggested as being preventative, or at least having the potential of slowing the pathological process [12]. The latest drug development strategies, in fact, target the pathogenetic mechanisms of Alzheimer's disease and, if successful, may be truly preventative [13].

Unfortunately, the value of potential preventative strategies remains as yet unrealized. It has been shown that risk factors identified from case–control studies do not necessarily translate into effective intervention studies, as in the case of folic acid supplementation and lowering of homocysteine for stroke prevention [14]. After the excitement related to the cognitive effects of oestrogen, the Women's Health Initiative Study found that oestrogen plus progestin therapy afforded no benefit to cognition in women aged 65 years and older [15]. In a recent study, high folate intake was paradoxically associated with cognitive decline in the elderly [16]. Drugs that slow the pathogenetic process may take many years to reach the clinic, as evidenced by the recent false start with an anti-Alzheimer's vaccine [17]. The control of cerebrovascular risk factors must start in midlife to make an impact, and brain reserve strategies are a lifelong commitment that should begin at school. Strategies on a mass scale and for prolonged periods will be difficult to examine empirically, and health budgets may not cope with interventions at such a scale. The prevention of dementia will force a re-think at many different levels of society.

Way forward with health policy

With regard to providing policy-makers with helpful information on the burden of dementia, we need to ensure that the concept of burden associated with the disease speaks the language of the policy-makers and can be delineated when needed. The 1996 Australian Burden of Disease statistics [18] showed that dementia ranked among the top six single diseases producing the greatest amount of burden. Currently, the metric for calculating burden of disease is in the form of the disability adjusted life year (DALY). This metric is made up of research indicating the relative risk of early death due to the disease (years of life lost) and the ‘disability weight’ due to the morbidity of the disease. The disability weight is a product of societal judgement on the non-desirability of living with that particular disorder. Interestingly, the disability weights used in Australian statistics are derived from a Dutch study [19]. Cost-effectiveness policy and health interventions programs are judged in relation to their ability to impact on ‘DALYs’. Therefore, it is time to ensure the disability weights in the DALY reflect Australian views on living with various stages of dementia. Our group anticipates the DALY will soon be updated, and it is important to ensure that useful data are collected for future burden of diseases estimates, both for dementia and predementia. Second, it is important to ensure that the information collected on disability will be easily transferred to a metric that is used in health policy-making. The latter issue is discussed here.

Currently, the DALY is viewed as a metric of burden of disease in terms of time lost in a healthy state. Dementia is a disease that predominantly affects older adults. With the current approach to DALY, economists could potentially generate a distorted view of the value of time lost between diseases that affect workers versus retirees, that is, the concept is ageist. For this reason, caution and ethical constraints are needed now and in the future when viewing DALYs solely as a time lost metric. Two courses of action may be helpful. First, we must document the DALYs for a predementia group that may have experienced disability in productive working years to ensure dementia begins to have a greater representation in the work of health economists involved in workforce issues. Second, if the DALY metric aids in planning and monitoring of intervention/public health programs, we must be involved in making the DALY a broader or dividable metric. The DALY subcomponents could be more relevant to dementia. Relevant subcomponents may include mobility, independence and life satisfaction, which are mutually exclusive to work productivity and so forth. In research, we do measure these factors, but we need to ensure in the future that this information can be transferred in a systematic way towards the DALY that the policy-makers use. Australian researchers have developed a measure that does such a thing (The Australian Quality of Life Instrument; AQoL) [20].

All health researchers need to consider how they can articulate their disability research into policy-friendly language and convert their information into potentially distinct DALY submetrics. Our group is using the AQoL in our longitudinal predementia study, and we look forward to making available the totalDALYs and subcomponent DALY scores for future Australian burden of disease publications. For those of us working in the field of documenting and defining prevalence and consequences of dementia disease stages, we must consider how our work will advance scientific knowledge as well as contribute to policy. Jorm et al.'s article in the issue shows us that even if partially effective interventions do come about in the future, the impact of the aging population alone will ensure that dementia is, and will be, a significant health issue for us all.

In Australia, dementia was named a health priority area this year (Australian Government 2005 Federal Budget: Ministry for Ageing Media Release JB120/05, 7 August 2005) [21]. Focus groups and forums are underway to establish research priorities, as well as to improve collaboration and sharing of information between groups [22]. The government's decision has provided the opportunity for us all to get involved now.