Article Text

The likely impact of demographic changes on the incidence and prevalence of neurological disease: demography in the United Kingdom
  1. G David Perkin
  1. Department of Clinical Neuroscience, Charing Cross Hospital, Fulham Palace Road, London W6 8RF, UK. Telephone 0181 846 1184

    Statistics from

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    The population of the United Kingdom, estimated to be 58.2 million in 1993, has been predicted to peak at 62.2 million by 2031,1 although recent projections have reduced that figure by at least a million.2 The recent reorganisation of The Health Service has emphasised the desirability of managing patients in the community and on an outpatient basis rather than in hospital. In England, for all specialties, the daily average number of available beds in 1983 was 343 000 compared with 220 000 in 1993-4.3 Over the same period, new outpatient attendance rose from 8 311 000 to 9 685 000. The falling number of beds has been offset by a substantial increase in throughput. In 1991-2 finished consultant episodes (defined as ordinary admissions combined with day cases) for diseases of the nervous system and sense organs (International Classification of Diseases-9th revision (ICD-9) codes 32-389) was 496 000, rising to 530 000 in 1993-4. This substantial increase in workload has been matched by increasing medical staffing levels. In England, all hospital neurological staff rose from 352 in 1983 to 446 in 1993, the consultant number rising, in the same period, from 146 to 186. The reverse process however, has occurred with nursing staff. Over the same period (1983 to 1993) their numbers fell from 374 851 to 344 503.3 By mid-1995, there were 250 consultant neurologists in England, Wales, Scotland, and Northern Ireland, amounting to one neurologist per 233 600 of the population.4

    The diseases encountered in neurological outpatient practice have been analysed both in surveys undertaken by individual neurologists,5 6 and in those conducted among a cohort of neurologists.7 8 In the survey undertaken by the association of British Neurologists,8 16 conditions were found to constitute 74% of the outpatient’s workload (table). Lists of this nature can, however, sometimes suggest a greater certainty of diagnosis than may actually exist. In one analysis of outpatient cases, a quarter were left without a specific diagnosis.6

    Top 16 conditions seen in neurology outpatient clinics

    Factors other than future demographic changes in the population are likely to influence the level and quality of healthcare provision. If neurologists, among their other outpatient responsibilities, were to see all stroke patients under the age of 65, and a third of those older, then one neurologist would be needed for every 98 000 of the population. If the planned expansion of trainees in neurology comes about, and assuming that those trainees achieve consultant status after five years of training, this ratio of neurologists to the population will not be reached till 2014.4 Changing patterns of management for certain conditions influence workload. It is now established, for example, that lone atrial fibrillation confers a fivefold increase in risk for ischaemic stroke.9 Regular review of such patients, particularly in the older age groups, will inevitably increase clinical work load. β-Interferon 1b is only the first of several drugs which can be shown to influence the natural history of multiple sclerosis. The cost of such drugs demands that their introduction is critically appraised and then carefully monitored. As an inevitable consequence, specialist clinics will be needed (and have been established), imposing an increased workload on a system already fully committed.

    The major demographic change in the United Kingdom over the next 30 years will be an increase in the proportion of elderly people. In 1993, 14.8% of the population were 65 or over, with 3.9% aged 80 or over. By 2031 the percentage aged 65 or over will be 22.9 and by 2051, the percentage aged 80 or over will be 9.2.1 The burden on health care can be represented in graphical form, showing projections for the number of dependent people (defined as men aged 65 or over, women aged 60 or over, and children under the age of 16) in the population in any given year. For older persons, the peak is reached about 2031 (fig 1). The problem, in terms of the financing of the health care of the very young and elderly, can be expressed as a dependency ratio (the combined number of children and pensioners for every 100 people of working age). The ratio was 63 in 1992 but is projected to rise to 82 in 2036.1

    Figure 1

    Projected dependent population in the United Kingdom until the year 2061.1

    Conditions which involve neurologists in their diagnosis and care, and which are likely to be influenced by this trend, include stroke and dementia.


    The incidence of stroke has fallen in recent years, as has stroke mortality (fig 2). A suggestion, from an analysis of the Rochester population, that that decline ceased in the 1980s in the United States, may be explicable on the basis of an increasing use of CT in patient assessment.10

    Figure 2

    Data from the Department of Health. The Health of the Nation, 1995.

    In The Netherlands, incidence rates for first stroke in males is predicted to reduce by 35% between 1985 and 2005.11 The trend is offset, however, by improving survival. In both The Netherlands and the United States, the average age of stroke patients is increasing. When calculations of survival are added to projected incidence rates, a trend towards increasing prevalence for stroke in very elderly people becomes apparent (fig 3). A projection for the population of England and Wales, which assumed a constant stroke incidence pattern in the future, produced a net increase in those who are moderately or severely handicapped by their first stroke six months after its onset of only 4% by 2023, despite a prediction of first ever strokes rising by 30% in the same period. The only modest increase in stroke prevalence predicted from this study, despite an aging population, was explained by the authors on their assumption that survival after first stroke, in very elderly people, was considerably less than in a younger cohort, particularly for those with more substantial disability.12

    Figure 3

    Graphs showing projected age specific absolute change in prevalence rate for major stroke in The Netherlands for the period 1985–2005. The middle curve (Rochester scenario) is based on incidence and case fatality data from Rochester, USA. Upper and lower limits have been calculated using the most extreme values of incidence and case fatality trends data from Niessen et al.11


    All epidemiological surveys have established that dementia increases with age and most suggest a higher incidence among women. Overall, dementia rates roughly double with every five years of aging. Most surveys have not attempted to differentiate between the different pathological subgroups, although postmortem data and data derived from clinical sources suggest that about two thirds of cases are due to Alzheimer’s disease.13 Using clinical criteria, the increase in incidence rates for dementia in elderly people seems to be due to cases of Alzheimer’s disease with little or no contribution from vascular dementia.14 Both incidence and prevalence studies of dementia have been performed. The first are disadvantaged by the need to recruit large samples and then follow them to allow serial examination. Reliance on cases being referred for assessment gives excessively low rates for incidence, with considerably higher figures being obtained from field studies.

    In a study of a United Kingdom population, annual incidence rates for subjects aged 75 to 79, 80 to 84, and 85 to 89 years were 2.3%, 4.6%, and 8.5% respectively. The study utilised the CAMDEX structured interview, and included only those with at least mild, as opposed to minimal dementia.15 If minimal cases were added, the respective quinquennial incidence rates became 5.1, 12.5, and 16.4.

    Substantially lower figures for incidence were obtained from another United Kingdom survey which had randomly recruited subjects in the relevant age groups from family practitioner lists.16

    The comparable incidence rates were 1.3%, 2.25%, and 2.18%. The surprisingly low figure for the 85-89 cohort is probably due to the small size of the sample. The screening procedures differed from the Cambridge survey, and the final assessment, using DSM-111-R criteria, included only those subjects with a dementing illness of at least moderate severity.

    A comparable study from the United States was based on the Framingham population.17 All new cases of dementia arising over a maximum of 10 years of follow up were ascertained. The incidence of dementia for the 65-69 age group and subsequent quinquennia were 0.7, 2.7, 5.2, 8.1, and 12%. The figures refer to moderate or severe cases and utilised NINCDS/ADRDA criteria.18 It was considered that Alzheimer’s disease accounted for rather more than half the cases.

    Based on 1989 Office of Population Censuses and Surveys figures, it is calculated that the total number of patients with dementia in the United Kingdom population aged 65 or over will have risen from 583 880 in 1991 to 735 480 in 2021. The impact of this increase both in terms of medical services and resource costs is substantial. A survey based on 1990/1 data has suggested that the cost of providing health and social services care to the population aged 65 or over with Alzheimer’s disease in England was £1039 million.19 If all cases of senile dementia are included, the cost would probably have reached £1500 million. By 2021, that cost, based on 1990/1 prices will have reached nearly £2000 million.


    The demographic change most liable to influence the healthcare burden over the next 20 to 30 years is the aging of the population. Despite the falling stroke incidence rate, a trend which may now have ceased, the increase in the elderly population will tend to increase the prevalence of stroke in those over 80, although that increase will be offset by the reduced likelihood of long term survival after first stroke in that age group. A major increase in cases of dementia in the population is predicted, with a corresponding rise in health costs. Factors other than demographic changes will continue to exert an influence on healthcare demands, including changes in the way in which the health service is structured, and developments in the treatment of certain conditions which lead to increased survival.