Epidemiology of motor neuron disease in Scotland, 1989-90 : a prospective study of incidence, clinical features and prognosis, and incorporating a case control study of antecedent environmental factors

Abstract

Chapter 1 The part that 19th century Edinburgh medicine played in formulating the early concepts of motor neuron disease as a nosological entity is described, including some hitherto unrecognized descriptions by graduates of Edinburgh Medical School, including the first description of progressive bulbar palsy. Credit is given to the great European neurologists, especially charcot’s landmark contribution of the definitive clinico-pathological correlations. Terminology relating to MND is discussed and the problems which have arisen over the nosology relating to diseases of motor neurons. The requirement for a clearly defined classification system and the importance of measuring incidence in this disease are outlined. Chapter 2 A comprehensive review of previous studies of incidence, mortality and distribution is presented. It is suggested that the value of previous clinical descriptions and epidemiological studies of motor neuron disease (MND) have been limited by methodological problems. There appears to have been a real increase in mortality from MND over recent decades. Comparison of incidence in different places is complicated by non standardised methods of case-ascertainment and diagnosis but evidence is presented which challenges the traditional concept that the distribution of MND in developed countries is uniform and static. It remains uncertain, from the evidence available, if age specific incidence continues to rise into old age and the apparent “peak” in late-middle life is due to ascertainment bias. This is an important consideration on which to base aetiological hypotheses. In the northern hemisphere there is a weak positive correlation between standardised, age specific incidence and distance from the equator. There is a high prevalence focus of an atypical MND/ALS on Guam, and while an environmental factor is probably responsible, its nature is uncertain. The relevance of other reported clusters of MND is discussed and the requirements for the ideal study of MND incidence are outlined, which acted as the impetus for the Scottish Motor Neuron Disease Register (SMNDR). Chapter 3 The methodology of the SMNDR is described. This is the first collaborative, population based, prospective, epidemiological study of MND and illustrates the feasibility of such a study and the way in which Scotland is suited (in terms of size and National Health service structure) to such a project. Diagnostic criteria are outlined for application to this and other large scale studies. Of 257 patients registered with the SMNDR as possible MND diagnosed in 1989 and 1990 in Scotland, 229 with proven (by autopsy), clinically definite or probable, sporadic or familial MND, form the basis of the description of incidence, distribution, clinical features and natural history contained in this chapter. The crude incidence was 2.24/100,000/year and age specific incidence continued to rise steeply with age into the very elderly age bands. No evidence for clustering on the basis of Scottish regions was found. 5% of patients had a family history of MND; the clinical pattern varied according to sex and age, with elderly women most likely to present with, or develop, progressive bulbar palsy. Chapter 4 The utility of the 1989-90 Scottish Hospital In-patient statistics (SHIPS) and 1989-90 death certificate coding (Registrar General for Scotland) by International classification of Diseases (ICD)-9, 335 (MND) are analysed as a tool for epidemiological studies and heath care planning. Coded hospital discharge data were an inaccurate record of a diagnosis of MND with a positive predictive value of a diagnosis of MND as determined by SHIPS of 70%. Such data cannot, in their present form, be used as a reliable measure of incidence in Scotland. Greater care is required in the preparation of summaries and coding if these data are to be useful care planning and epidemiological research. However, as an important source of case notification for the SMNDR to achieve a complete sample of patients. There was also a problematic false positive rate (10%) for mortality data but this source more closely approximated true incidence. Chapter 5 A review of previous case control studies of environmental risks for MND is presented. In order to test the hypothesis that certain environmental factors may play a role in the aetiology of MND, a case control study of 103 incident patients (diagnosis date May 1990-october 1991; 57 from the above cohort (as described in chapter 3) and 46 diagnosed in 1991, age and sex matched with community controls was conducted and is described. Measures were taken to minimise potential sources of bias and these are discussed. Fractures were more common in patients than controls especially in the in the five years prior to symptom onset (odds ratio = 15 (95% CI, 2.3-654). Manual workers were over represented and a number of environmental exposures of potential toxicity including exposure to lead (OR = 5.3, 95% CI, 1.5-11) and solvents or chemicals (OR = 3.8, 95% CI 1.5-11) were found. The limited extent of these associations favours a multifactorial aetiology for MND. No relationship to social class, poliovirus infection or to factors which might increase the risk of enteroviral infection in childhood (home space and domestic amenities) was found. Chapter 6 A comprehensive review of previous prognostic studies in MND is presented. An actuarial analysis of the survival of the cohort of 229 incident patients (57 seen in person) in Scotland in 1989-90, with complete follow up for a mean of two years from diagnosis, is calculated on the basis of the SMNDR clinical classification system. The overall 50% survival from diagnosis was 1.2 years (95% CI, 1.0-1.4) and from symptom onset, 2.5 years (95% CI, 2.2-3.0 years). There were significant differences in survival with a poorer prognosis for: progressive bulbar palsy (PBP), female sex and age > 65 years. The presence of PBP and old age were the strongest predictors of outcome, the difference in survival between the sexes was due to the higher frequency of PBP in females. The overall five year survival from symptom onset was 27.7% (95% CI, 19.9-36.0%) but for patients with PBP as the presenting feature was only 3.5% (95% CI, 0.0-15%). These results suggest that, in well defined cohorts, survival can be predicted with some confidence and are useful for the planning of treatment trials. Chapter 7 In this concluding chapter I consider the work contained in this thesis.