Deaths Due to Dementia:
An Analysis of Multiple-cause-of-death Data

Kathryn Wilkins, Greg F Parsons, Jane F Gentleman and William F Forbes

Abstract

This study analyzes multiple-cause-of-death information from over 113,000 death certificates of Canadians aged 65+ and identifies causes that are significantly likely and significantly unlikely to combine with dementia to cause death. For dementia as a mentioned cause and as the underlying cause of death, frequencies and rates of death were calculated. Dementia was mentioned on death certificates 2.4 times as often as it occurred as the underlying cause of death. Among the causes least associated with dementia were some cancers, chronic respiratory diseases and rheumatoid arthritis. Causes of death that rarely occur with dementia should be further investigated in terms of their potential role in preventing or delaying the onset of dementia. In particular, further study of the role of anti-inflammatory drugs and nicotine in reducing the risk of dementia is indicated. Causes positively associated with dementia largely reflect the physical deterioration it confers.

Key words: aging; Canada; cause of death; death certificate; dementia

Introduction

With the aging of the Canadian population, the number of people affected by dementia is increasing substantially. The prevalence of dementia in Canadians aged 65 and over has been estimated at 8%, amounting to over 250,000 people.¹ The economic cost of dementia in Canada is estimated to be over $3.9 billion annually, or about $14,000 per patient per year.²

Athough the significance of dementia in terms of its prevalence, debilitating effects and caretaker burden is increasingly appreciated, its contribution to causing death is less well understood. Dementia, characterized by confusion, disorientation and intellectual impairment, would not by itself seem to be a life-threatening disorder. However, as it advances, dementia gives rise to complications that result in loss of physical function, and thus it is the consequences of the initial disease that cause death.

 When dementia is the disorder that has initiated the sequence of events leading to death, it is considered to be the underlying cause of death, and, according to the international conventions that govern the medical certification of death, it is entered as such on the death certificate.³ Dementia might also be mentioned (but not as the underlying cause) on the death certificate if it is considered to have "unfavorably influenced the course of the morbid process, and thus contributed to the fatal outcome," even though it was "not related to the disease or condition directly causing death."³ A dementia code can thus be given as either the underlying cause or a non-underlying cause. However, for a person with dementia whose death is considered to have been caused by other conditions or circumstances, dementia will not necessarily be included on the death certificate.

 Conventionally, published mortality statistics identify only a single, underlying condition as the cause of death. However, this practice results in a loss of information, because most deaths result from several disorders, especially in the elderly. By virtue of their age, people with dementia are often afflicted with other diseases as well. Causes of death that appear on the death certificate with dementia may be complications of dementia, or they may reflect other pathology. The analysis of multiple-cause-of-death data, which include all causes of death entered on the death certificate, permits a more accurate characterization of death than does analysis of single-cause data.^3-5 The availability of such data in Canada is increasing, although few analyses have yet been published.

In countries other than Canada, studies of multiple-cause-of-death data have been reported for some time, and a number of them have focused on dementia. For example, multiple-cause data have been used to illustrate the deficit in the frequency with which dementia is certified as a cause of death, as compared with its presence as a cause of morbidity before death.^6-13 Other analyses of multiple-cause data have focused on the causes of death that are positively associated with dementia.^14,15

 The purpose of this paper is to extend current knowledge of the role of dementia as a reported cause of death using multiple-cause-of-death data from Canada and the United States. Comparisons are made between the frequencies with which dementia is certified as the underlying cause of death and as a mentioned cause. Using Canadian data, we report the causes of death that are positively linked to dementia, as well as those that are negatively associated.

Methods

The development of the data for analysis has been described previously.⁵ Briefly, multiple-cause records (produced by automated processing of medical information from death certificates) were edited using software designed for this purpose,¹⁶ and then linked with death records for corresponding individuals from the Canadian Vital Statistics Data Base maintained by Statistics Canada. Thus, data on age at death, sex and other variables were added to each record on the multiple-cause file.

 After further edits, 113,144 records of decedents aged 65 and over were retained for analysis. These represented 19% of all deaths of people aged 65 and over  in Canada from 1990 to 1993. For the provinces of Quebec and Ontario, random samples of 3% of all death records for the years 1992 and 1993, respectively, were analyzed. For Alberta, New Brunswick, Newfoundland, Nova Scotia, Prince Edward Island, Saskatchewan, Northwest Territories and Yukon, nearly complete data were used. Multiple-cause data were not available for British Columbia or Manitoba. Complete 1993 multiple-cause-of-death data for the United States, available on CD-ROM from the National Center for Health Statistics,¹⁷ had undergone the same type of processing as the Canadian data.

 The integrity of the Canadian multiple-cause file was verified using comparisons with other sources. The percentage distribution of the age at death on the records in the multiple-cause file was compared with the same distribution for all deaths in Canada in 1991. As well, the median ages at death and the percentage of deaths with dementia as the underlying cause for age groups 45+, 75+ and 45S74 in the multiple-cause records were compared with the corresponding results reported for all deaths.¹⁸ These comparisons showed only negligible differences.

 Dementia was defined in this study as either senile and presenile organic psychotic conditions (ICD-9 code 290) or Alzheimer's disease (ICD-9 code 331.0). Frequencies, rates and proportions of all Canadian and US deaths for which dementia was mentioned as a cause were tabulated by sex and age group. For the Canadian data, calculations by sex and age were made for rates of dementia (as a mentioned cause and as the underlying cause of death), mentioned-to-underlying-cause ratios (i.e. the number of times dementia was mentioned at all on the death certificates divided by the number of times it was selected as the underlying cause of death) and rate ratios. Age-standardized rates were calculated by the direct method using the 1991 Canadian population counts in the age groups 65-69, 70-74, 75-79, 80-84, 85-89 and 90+.

 To estimate bivariate associations between dementia and every other cause on the same death certificate, odds ratios (ORs) were calculated by sex and age group. These are the odds of dementia being mentioned given that another specific cause is mentioned, divided by the odds of dementia being mentioned given that the other cause is not mentioned. ORs were calculated for all causes at the level of the first three digits in the ninth revision of the International Classification of Diseases (ICD-9).³ For the mentioned-to-underlying-cause ratios and ORs, dementia was counted at most once per death certificate, as was any other cause. Note, however, that when a code is entered on a death certificate to indicate the nature of an injury that is a cause of death (i.e. the N-code), a corresponding code is always entered to indicate the external cause of the injury (i.e. the E-code). For the calculations of ORs, both N- and E-codes were included from each record containing such entries.

 Two-tailed tests at the 0.05 significance level were used to determine whether ORs were significantly different from 1.00. This is equivalent to testing for independence in a two-by-two contingency table of frequencies of dementia being mentioned or not mentioned on the death record, versus the other cause being mentioned or not mentioned. To increase the likelihood of clinical significance of observed associations, only ORs for which each frequency in the two-by-two table was at least 10 were analyzed. The large samples conferred high power on the significance tests, allowing the tests to detect a large number of significant results.

Results

Dementia was given as the underlying cause of death in 2% of all the Canadian death records of people aged 65 and over, and it was mentioned in 6% of these records. Although the pattern of the US data was strikingly similar, the proportions of Canadian death records that mentioned dementia were consistently slightly higher than in the US data (Tables 1, 2; Figures 1, 2). As well, the mentioned-to-underlying-cause ratio was slightly lower for the Canadian death certificates than for the US death certificates-2.4 and 2.5, respectively.

Death rates for dementia increased sharply with age, irrespective of whether rates were based on the underlying or mentioned cause of death (Table 3, Figure 1). Among Canadians aged 85 and over, the rate of death for which dementia was the underlying cause (525 per 100,000 population) was about 26 times as high as among those aged 65S74 (20 per 100,000 population). This compares with an approximately 32-fold increase in the corresponding age groups (1348 and 42 per 100,000 population respectively) when dementia was mentioned on the death certificate. In comparison, the 1992 rate of all-cause mortality for the age group 85+ was only seven times as high as the rate for the age group 65-74.¹⁹ That is, dementia death rates increase with age more rapidly than all-cause mortality rates.

 The proportion of death certificates giving dementia either as the underlying cause or as a mentioned cause was higher among women than men (Tables 1 and 2, Figures 1 and 2). The age-standardized rate of death due to dementia, however, was higher among men (Table 3).

 To identify causes that were positively or negatively associated with dementia on death certificates, odds ratios were tested for significance (see Methods). A total of 153 tests were conducted for all ages combined (65+), both sexes and all three-digit ICD codes. Of these tests, 70% were significant at the 0.05 level, which overwhelmingly exceeds the figure of 5% that would be expected due to chance alone. Similarly, 231 tests were conducted for three separate age groups (65-74, 75S84 and 85+), of which 56% were significant.

Causes Positively Associated with Dementia

For 36 causes of death, the ORs were significantly elevated in at least one age group, for at least one of the sexes (Table 4). The causes of death that were positively associated with dementia fell into several general categories, as follows: pneumonia and influenza; conditions that are symptomatic of dementia; conditions that arise from the debilitating effects of advanced dementia (e.g. difficulty eating, incontinence, immobility); and cerebrovascular disease. Within these categories, the specific causes of death for which ORs were significantly elevated for both men and women aged 65 and over are shown in Table 5. The causes of death with the most consistently elevated ORs over the age groups and within the two sexes included bronchopneumonia, organism unspecified (ICD-9 485); pneumonia, organism unspecified (ICD-9 486); and pneumonitis due to solids and liquids (ICD-9 507).

Conditions Negatively Associated with Dementia

Forty-one causes of death were significantly negatively associated with dementia (Table 6). These causes could be generally categorized as cancer, heart disease, chronic respiratory diseases and other causes. Within these groupings, the specific causes of death for which the ORs were significantly low for both males and females in the age group 65+ are shown in Table 7. Consistently low ORs were noted for acute myocardial infarction (ICD-9 410), other forms of chronic ischemic heart disease (ICD-9 414) and cardiac dysrhythmias (ICD-9 427). Although relatively rarely certified as a cause of death, rheumatoid arthritis and other inflammatory polyarthropathies (ICD-9 code 714) achieved statistical significance (OR = 0.32) among women in the age group 65+. Among men, the OR for this cause of death was 0.39, significant at level 0.06.

TABLE 6 Significant negative associations between mention of dementia and mention of other causes of death, by sex and age group, ages 65+, Canada,a 1990-1993
		Odds ratio
		Men, by age group				Women, by age group
ICD-9 code	Other causesb	65-74	75-84	85+	65+	65-74	75-84	85+	65+
153	Malignant neoplasm of colon	0.24	0.46*	0.23	0.33*	0.38	0.28*	0.29*	0.27*
159	Malignant neoplasm of other and ill-defined sites within the digestive organs and peritoneum	-	0.62	0.09	0.34*	0.19	0.24	0.23	0.21*
162	Malignant neoplasm of  trachea, bronchus and lung	0.24*	0.22*	0.20	0.17*	0.16	0.19*	0.38	0.15*
174	Malignant neoplasm of female breast	N/A	N/A	N/A	N/A	0.33	0.47*	0.68*	0.42*
185	Malignant neoplasm of prostate	0.34	0.34*	0.58*	0.49*	N/A	N/A	N/A	N/A
188	Malignant neoplasm of bladder	0.29	0.34	0.57	0.45*	-	0.66	0.09	0.32
197	Secondary malignant neoplasm of respiratory and digestive systems	0.18	0.12	0.13	0.11	-	0.13	0.21	0.10
198	Secondary malignant neoplasm of other specified sites	0.90	0.18	0.10	0.11	-	0.24	0.33	0.14
199	Malignant neoplasm without specification of site	0.11	0.15*	0.22*	0.14*	0.21	0.17*	0.22*	0.15*
250	Diabetes mellitus	1.15	0.96	0.80	0.89	0.72	0.59	0.65	0.59*
285	Other and unspecified anemias	0.38	0.59	0.92	0.90	0.38	0.69	0.55*	0.66*
303	Alcohol dependence syndrome	1.24	0.59	1.22	0.56*	0.59	-	2.24	0.27
310	Specific non-psychotic mental disorders following organic brain damage	0.94	0.79	0.26	0.63	0.85	0.55	0.28*	0.44*
401	Essential hypertension	0.96	0.60*	0.79	0.67*	0.54	0.66*	0.73*	0.67*
402	Hypertensive heart disease	1.15	0.78	0.98	0.86	0.53	0.39	0.76	0.64*
403	Hypertensive renal disease	0.62	0.56	0.70	0.63	0.77	0.32	0.20	0.29*
410	Acute myocardial infarction	0.40*	0.40*	0.51*	0.40*	0.27*	0.33*	0.43*	0.35*
414	Other forms of chronic ischemic heart disease	0.53*	0.57*	0.63*	0.59*	0.55*	0.49*	0.61*	0.60*
415	Acute pulmonary heart disease	0.42	0.29	0.44	0.33*	0.16	0.86	0.44*	0.54*
424	Other diseases of endocardium	-	0.19	0.69	0.35*	0.36	0.35	0.37*	0.38*
427	Cardiac dysrhythmias	0.67*	0.79*	0.84*	0.81*	0.90	0.84*	0.77*	0.85*
428	Heart failure	0.71	0.60*	0.61*	0.74*	0.73	0.40*	0.50*	0.56*
429	Ill-defined descriptions, complications of heart disease	0.14	0.53*	0.55*	0.47*	1.14	0.80	0.95	0.92
431	Intracerebral hemorrhage	0.32	0.24	0.93	0.41*	-	0.66	0.45	0.45*
434	Occlusion of cerebral arteries	0.98	0.63	0.69	0.73	-	1.08	0.42*	0.68*
436	Acute but ill-defined cerebrovascular disease	2.08*	1.17	0.72*	1.18*	1.91*	0.88	0.67*	0.89*
441	Aortic aneurysm	-	0.29	0.06	0.17*	0.25	0.38	0.74	0.52*
492	Emphysema	0.36	0.61*	0.28	0.46*	0.27	0.10	0.13	0.11*
493	Asthma	-	0.53	1.45	0.73	-	0.54	0.64	0.49*
496	Chronic airways obstruction, not elsewhere specified	0.84	0.55*	0.60*	0.62*	0.59	0.42*	0.56*	0.45*
518	Other diseases of lung	0.27	0.80	0.38	0.57*	0.26	0.75	0.66	0.61*
557	Vascular insufficiency of intestine	-	0.12	0.33	0.18	0.34	0.51	0.45	0.46*
560	Intestinal obstruction without mention of hernia	1.28	0.51	0.65	0.70	0.53	0.64	0.74	0.70*
584	Acute renal failure	-	0.89	0.75	0.75	0.35	0.34	0.46	0.38*
585	Chronic renal failure	-	0.43*	0.63	0.54*	0.70	0.23	0.50*	0.38*
586	Renal failure, unspecified	0.61	0.74*	0.45*	0.66*	0.16	0.24*	0.58*	0.43*
714	Rheumatoid arthritis and other inflammatory polyarthropathies	0.76	0.36	0.40	0.39	0.27	0.25	0.50	0.32*
785	Symptoms involving cardiovascular system	-	0.31*	0.96	0.47*	0.63	0.44*	0.73	0.57*
797	Senility without mention of psychosis	5.33	1.39	0.72*	1.40*	7.50	1.46*	0.56*	0.95
N-997	Complications affecting specified body systems, not elsewhere specified	-	0.61	0.38	0.46	0.47	0.51	0.61	0.51*
E-878	Surgical operation and other surgical procedures as the cause of abnormal reaction of patient, or of later complication, without mention of misadventure at the time of operation	0.24	0.48	0.45	0.41*	0.24	0.37	0.43	0.34*
a Records were not available for all jurisdictions. See Methods. b Includes all 3-digit ICD-9 codes with a significant odds ratio greater than 1.00 in at least one age group and with frequencies of at least 10 in each cell of the 2-by-2 table. * Odds ratio is significant. Significance tests were two-tailed and at level 0.05. N/A: Not applicable in cases of sex-specific conditions

Discussion

This study, based on analyses of multiple-cause-of-death data, reveals new information about the total frequency with which dementia is certified as a cause of death, as well as the causes of death that most frequently and most infrequently occur with dementia. The benefit of multiple-cause data compared with conventional single-cause mortality data is particularly pertinent in the study of dementia, for which a full epidemiologic understanding has not yet been gained.

The finding that a higher proportion of Canadian than US records mentioned dementia might reflect one or both of the following: 1) a more comprehensive identification of dementia cases before death in Canada than in the US, perhaps because of universal access to free diagnostic services and medical care in Canada; 2) a higher proportion of deaths caused by dementia in Canada than in the US.

The increase with age of the mentioned-to-underlying-cause ratio in both countries suggests that when dementia is entered on the death certificate at lower ages, it is more likely to be recorded as the underlying cause of death than when it is entered at higher ages. This may reflect the assumption that dementia in old age is "normal," and thus less likely to initiate the sequence of events leading to death. Increases with age of dementia as a mentioned cause of death can be compared with increases with age in the reported prevalence of dementia in the population. For example, according to the results of a population-based prevalence study, dementia (of all types) was estimated to occur in 2.4% of the Canadian population aged 65S74, rising to 34.5% of those aged 85 and over, an increase by a factor of about 15.¹ In comparison, the death rate based on the mention of dementia increases by a factor of 32 between the two age groups. This difference in the two ratios occurs presumably because people aged 65S74 who have dementia are less likely to die from it than are people at older ages.

The higher death rates among men than women contrast with prevalence estimates showing a higher rate of dementia in women.¹ This difference probably reflects the longer survival time of women with dementia, for which there is ample evidence.^13,14,20,21

Beginning at about the age of 86, however, the death rates among women surpass those among men (Figures 1 and 2). A variety of reasons could account for this crossover. Although people of both sexes over age 85 represent a very select population, men at that age may be generally healthier than women if a greater proportion of less healthy men have previously been removed from the cohort by death. Or, at these highest ages, women who have a longer life expectancy are more likely to develop "severe" dementia, which is thus more likely to be recorded on death certificates for women than for men.

One explanation of the consistently higher mentioned-to-underlying-cause ratio among men than women is that co-morbid conditions such as coronary heart disease are more often lethal in men than women, even though women generally have more reported co-morbid conditions (unpublished data, National Population Health Survey, 1994/95). Causes of death other than dementia would thus be more frequently designated as the underlying cause among men than among women.

Positively Associated Causes

The strong, consistent association of dementia with bronchopneumonia, pneumonia and influenza has been observed previously¹⁴ and probably reflects the opportunistic nature of these illnesses. People whose general health is compromised by the physical effects of dementia are presumably less resistant to ailments that might otherwise not result in death. As well, pharmaceutical and other treatments for these conditions might be prescribed with less stringency to people with dementia than to those not so affected.

Some causes of death positively associated with dementia on the death certificate arose as part of the symptomatology of dementia. Depression, for example, may be the presenting feature or an early symptom of dementia. Similarly, other causes of death positively associated with dementia reflect the physiological deterioration, such as difficulty eating, incontinence, motor dysfunction and eventual immobility, that occurs as dementia progresses. For example, difficulty with eating could result in malnutrition, disorders of fluid, electrolyte and acidSbase balance, pneumonitis and symptoms involving the digestive system (e.g. difficulty in swallowing).

The results of the analysis pertaining to cerebrovascular as well as cardiovascular diseases (see discussion below) should be interpreted with caution. Because the death certificate often contains the more general "dementia" diagnosis rather than a specific subtype, all elements of senile and presenile organic psychotic conditions, together with Alzheimer's disease, were treated as one disease in the analysis. However, it is likely that associations with other disorders are not constant across subtypes of dementia. For example, arteriosclerotic dementia and Alzheimer's disease are likely to have quite different associations with stroke and hypertension-related diseases.

Negatively Associated Causes

The negative associations observed between dementia and causes of death for which smoking is a risk factor, particularly cancers of the trachea, bronchus and lung, and chronic respiratory diseases, are consistent with the hypothesis that nicotine can reduce the development of Alzheimer's disease. The negative associations with respiratory cancers might also be partially explained by their rapidly fatal course, however, which may lower the likelihood of certification of dementia as a cause of death.

Although the evidence is equivocal, some epidemiologic studies suggest that nicotine may protect against Alzheimer's disease^22-25 by increasing the number of nicotinic cholinergic receptors.^26-29 In addition, nicotine may enhance cognition in normal individuals,³⁰ in patients who already have Alzheimer's disease^31-32 and in animals.³³

Findings from the multiple-cause-of-death data also indicate a negative association between dementia and heart disease, another cause of death for which smoking is a risk factor. Although previous research has suggested a physiologic association between the formation of senile plaques in the brain and coronary artery disease,³⁴ the negative association observed in the present study is consistent with the results of a study of deaths among dementia patients in Finland: among these patients, cardiovascular disease was less often certified as a cause of death than in the general population.¹⁴ However, as with lung cancer, the rapid and dramatic course of some types of heart disease might partially account for the negative association observed with dementia. For example, dementia might be certified less frequently for patients who die suddenly from an acute myocardial infarction than from other, more lingering causes.

The negative association observed between essential hypertension and dementia corroborates the findings of other research showing a lower prevalence of Alzheimer's disease among people with hypertension than among those without.³⁵ Dementia was also negatively associated with other causes of death for which hypertension is a risk factor, including aortic aneurysm and kidney failure.

A recent review of the literature indicates that at least 20 studies suggest that the use of anti-inflammatory drugs for the treatment of rheumatoid arthritis is associated with a lower prevalence of Alzheimer's disease.³⁶ The significantly low OR observed in the present study for rheumatoid arthritis and other inflammatory polyarthropathies (ICD-9 714) among women aged 65 and over is consistent with the hypothesis that anti-inflammatory drugs can delay or prevent the onset of Alzheimer's disease.

Limitations

In interpreting the results of the analysis of multiple-cause data, one must be mindful of the data's limitations.⁴ First, certification practices may affect some of the observed associations between causes of death and dementia. For example, previous research suggests that, for people with dementia, the likelihood of certification of particular causes of death, including Alzheimer's disease, pneumonia, heart disease and stroke, varies with the level of the patient's cognitive impairment before death.¹⁵

It is also important to note that mortality statistics, including multiple-cause-of-death data, do not fully reflect disease prevalence in those who died. This is because the causes that a person dies from do not necessarily include all the diseases that he or she dies with. As well, a decedent's medical history may not be fully known to the authority who certifies the death, and so a condition that was involved in the sequence of events leading to death may not be entered on the death certificate. Clearly, the greatest limitation to the present study is the potential for misclassification bias. If undercertification of dementia is randomly distributed relative to other causes of death, then the effect will be to decrease the magnitude of the observed associations. More likely, however, certification of dementia relates somewhat to the presence of specific other causes of death, which will result in some amount of differential misclassification and error in the results.

Nonetheless, the analysis of multiple-cause-of-death data provides added insight into the particular combinations of conditions that are fatal, as well as conditions that rarely combine to cause death. Research done to date, mostly involving case-control comparisons, has revealed relatively little regarding the etiology of the major dementias. Multiple-cause data provide a low-cost means of studying cases in which dementia causes death, and also the causes of death with which dementia occurs most rarely.

In conclusion, multiple-cause-of-death data reveal that dementia contributes to death more than twice as often as it is identified as the underlying cause of death. Many of the causes of death that are positively linked to dementia on the death certificate are attributable, directly or indirectly, to the dementing illness, a finding that is consistent with current knowledge. Causes that are negatively associated with dementia may involve treatments that help to prevent or delay the onset of dementia. For example, the negative associations observed between dementia and smoking-related cancers and arthritis are supportive evidence of possible protective effects of nicotine and anti-inflammatory drugs. Further study of the roles of these substances in preventing or delaying the onset of dementia is indicated. Other negative associations observed, such as those between dementia and hypertension, aortic aneurysm and renal failure, suggest a variety of hypotheses that warrant further investigation.

References

1. Canadian Study of Health and Aging Working Group. Canadian Study of Health and Aging: study methods and prevalence of dementia. Can Med Assoc J 1994;150(6):899-913.

2. Østbye T, Crosse E. Net economic costs of dementia in Canada. Can Med Assoc J 1994;151(10):1457-64.

3. World Health Organization. International Classification of Diseases. Manual of the international statistical classification of diseases, injuries, and causes of death. 1975 [ninth] revision. Geneva: WHO, 1977.

4. Israel RA, Rosenberg HM, Curtin LR. Analytical potential for multiple cause-of-death data. Am J Epidemiol 1986;124(2):161-79.

5. Wilkins K, Wysocki M, Morin C, et al. Multiple causes of death. Health Reports 1997;9(2):19-29. (Statistics Canada Cat 82-003-XPB).

6. Martyn CN, Pippard EC. Usefulness of mortality data in determining the geography and time trends of dementia. J Epidemiol Community Health 1988;42:134-7.

7. Burns A, Jacoby R, Luthert P, et al. Cause of death in Alzheimer's disease. Ageing 1990;19:341-4.

8. Morgan K, Lilley J, Arie T, et al. Incidence of dementia: preliminary findings from the Nottingham longitudinal study of activity and ageing. Neuroepidemiology 1992;11(Suppl 1):80-3.

9. Macera CA, Sun RKP, Yeager KK, et al. Sensitivity and specificity of death certificate diagnoses for dementing illnesses, 1988-1990. J Am Geriatr Soc 1992;40:479-81.

10. Newens AJ, Forster DP, Kay DWK. Death certification after a diagnosis of presenile dementia. J Epidemiol Community Health 1993;47:293-7.

11. Raiford K, Anton-Johnson S, Haycox Z, et al. CERAD part VII: Accuracy of reporting dementia on death certificates of patients with Alzheimer's disease. Neurology 1994;44:2208-9.

12. Olichney JM, Hofstetter CR, Galasko D, et al. Death certificate reporting of dementia and mortality in an Alzheimer's disease research center cohort. J Am Geriatr Soc 1995;43:890-3.

13. Jagger C, Clarke M, Stone A. Predictors of survival with Alzheimer's disease: a community-based study. Psychol Med 1995;25:171-7.

14. Mölsä PK, Marttila RJ, Rinne UK. Survival and cause of death in Alzheimer's disease and multi-infarct dementia. Acta Neurol Scand 1986;74:103-7.

15. Kukull WA, Brenner DE, Speck CE, et al. Causes of death associated with Alzheimer's disease: variation by level of cognitive impairment before death. J Am Geriatr Soc 1994;42:723-6.

16. Chamblee RF, Evans MC. TRANSAX: the NCHS system for producing multiple-cause-of-death statistics, 1968-78. Washington (DC): US Government Printing Office, 1986 Jun; Public Health Service, Vital and Health Statistics, Series 1, No 20.

17. National Center for Health Statistics. 1993 multiple cause-of-death file [CD-ROM Series 20, No 11]. US Department of Health and Human Services, 1997 Jul.

18. Forbes WF, Gentleman JF, Park E. Dementia and age at death. Exp Gerontol 1995;30:445-53.

19. Statistics Canada. Causes of death, 1992. Ottawa, 1994; Cat 84-208.

20. van Dijk PTM, van de Sande JH, Diederik WJD, et al. The nature of excess mortality in nursing home patients with dementia. J Gerontol 1992;47:M28-34.

21. van Dijk PTM, Dippel DWJ, van der Meulen JHP, et al. Comorbidity and its effect on mortality in nursing home patients with dementia. J Nerv Ment Dis 1996;184(3):180-7.

22. Graves AB, van Duijn DM, Chandra V, et al. Alcohol and tobacco consumption as risk factors for Alzheimer's disease: a collaborative re-analysis of case-control studies. Int J Epidemiol 1991;20:S48-57.

23. Forbes WF, Hayward LM. Concerning the link between tobacco smoking and impaired mental functioning. Can J Public Health 1993;84(3):211-2.

24. van Duijn CM, Hofman A. Relation between nicotine intake and Alzheimer's disease. Br Med J 1991;302:1491-4.

25. Lee PN. Smoking and Alzheimer's disease: a review of the epidemiological evidence. Neuroepidemiology 1994;13(4):131-44.

26. Perry EK, Perry RH, Smith CJ, et al. Nicotinic receptor abnormalities in Alzheimer's and Parkinson's diseases. J Neurol Neurosurg Psychiatry 1987;50:806-9.

27. Whitehouse PJ, Martino AM, Marcus KA, et al. Reductions in acetylcholine and nicotine binding in several degenerative diseases. Arch Neurol 1988;45:722-4.

28. Whitehouse PJ, Martino AM, Wagster MV, et al. Reductions in [3H] nicotinic acetylcholine binding in Alzheimer's disease and Parkinson's disease: an autoradiographic study. Neurology 1988;38:720-3.

29. Whitehouse PJ, Kalaria RN. Nicotinic receptors and neurodegenerative dementing diseases: basic research and clinical implications. Alzheimer Dis Assoc Disord 1995;9 (Suppl 2):3-5.

30. Warburton DM. Nicotine as a cognitive enhancer. Prog Neuropsychopharmacol Biol Psychiatry 1992;16:181-91.

31. Newhouse P, Sunderland T, Tariot P, et al. Intravenous nicotine in Alzheimer's disease: a pilot study. Psychopharmacology 1988;95:171-5.

32. Sahakian B, Jones G, Levy R, et al. The effects of nicotine on attention, information processing, and short term memory in patients with dementia of the Alzheimer type. Br J Psychiatry 1989;154:797-800.

33. Williams M, Sullivan JP, Arneric SP. Neuronal nicotinic acetylcholine receptors. Drug News Perspectives 1994;17:205-0.

34. Sparks DL, Hunsaker JC III, Scheff SW, et al. Cortical senile plaques in coronary artery disease, aging and Alzheimer's disease. Neurobiol Aging 1990;11(6):601-7.

35. Rockwood K, Lindsay J, McDowell I. High blood pressure and dementia [letter]. Lancet 1996;348:65.

36. McGeer EG, McGeer PL. Aging, neurodegenerative disease and the brain. Can J Aging 1997;16(2):218-36. 

Author References
Kathryn Wilkins, Health Statistics Division, Statistics Canada, Tunney's Pasture, 18C - RH Coats Building, Ottawa, Ontario  K1A 0T6
Greg F Parsons, Distributive Trades Division, Statistics Canada, Ottawa, Ontario
Jane F Gentleman, Social Survey Methods Division, Statistics Canada, Ottawa, Ontario
William F Forbes, Elisabeth-Bruyère Health Centre, Ottawa, Ontario

[Previous][Table of Contents] [Next]