Background Reported outcomes for colorectal cancer associated with Inflammatory Bowel Disease are inconsistent. We compared survival outcomes in colorectal cancer patients with and without Inflammatory Bowel Disease using a population-based cohort and elicited prognostic factors associated with survival Methods Adult patients with a diagnosis of colorectal cancer in 2007-2015 were identified from the Ontario Cancer Registry. Those with Inflammatory Bowel Disease were detected via the validated Ontario Crohn’s and Colitis Cohort. Primary outcome measure was overall survival from time of colorectal cancer diagnosis until the date of death. Secondary outcome measures included treatments received and publicly-provided health care costs. Results Colorectal cancer was diagnosed in 67,137 with Inflammatory Bowel Disease present in 783 (1.2%). The Inflammatory Bowel Disease-associated colorectal cancer patients were younger at diagnosis (median range 55-59 vs 70-74, p<0.001). Five-year survival in Inflammatory Bowel Disease-associated patients was 56.4% (95% CI 52.6-59.9) and 57.0% (95% CI 56.6-57.4) in sporadic colorectal cancer (p=0.8). Inflammatory Bowel Disease was a significant predictor of death (Hazard Ratio=1.45, 95% CI 1.29-1.63, p<0.001) after adjusting for other variables. In patients under 50, 5-year survival was significantly (p<0.001) reduced in the Inflammatory Bowel Disease population (56.8%, 95% CI 49.4-63.5) compared with the sporadic colorectal cancer population (71.4%, 95% CI 70.0-72.7). Similar results were observed in those 50-64 years old. Conclusion Young patients (<65) with Inflammatory Bowel Disease-associated colorectal cancer have worse survival outcomes than young (<65) patients with sporadic colorectal cancer. These findings inform prognostication and may direct future research for this high-risk population. / Thesis / Master of Science (MSc) / Background Reported outcomes for colorectal cancer associated with Inflammatory Bowel Disease are inconsistent. We compared survival outcomes in colorectal cancer patients with and without Inflammatory Bowel Disease using a population-based cohort and elicited prognostic factors associated with survival Methods Adult patients with a diagnosis of colorectal cancer in 2007-2015 were identified from the Ontario Cancer Registry. Those with Inflammatory Bowel Disease were detected via the validated Ontario Crohn’s and Colitis Cohort. Primary outcome measure was overall survival from time of colorectal cancer diagnosis until the date of death. Secondary outcome measures included treatments received and publicly-provided health care costs. Results Colorectal cancer was diagnosed in 67,137 with Inflammatory Bowel Disease present in 783 (1.2%). The Inflammatory Bowel Disease-associated colorectal cancer patients were younger at diagnosis (median range 55-59 vs 70-74, p<0.001). Five-year survival in Inflammatory Bowel Disease-associated patients was 56.4% (95% CI 52.6-59.9) and 57.0% (95% CI 56.6-57.4) in sporadic colorectal cancer (p=0.8). Inflammatory Bowel Disease was a significant predictor of death (Hazard Ratio=1.45, 95% CI 1.29-1.63, p<0.001) after adjusting for other variables. In patients under 50, 5-year survival was significantly (p<0.001) reduced in the Inflammatory Bowel Disease population (56.8%, 95% CI 49.4-63.5) compared with the sporadic colorectal cancer population (71.4%, 95% CI 70.0-72.7). Similar results were observed in those 50-64 years old. Conclusion Young patients (<65) with Inflammatory Bowel Disease-associated colorectal cancer have worse survival outcomes than young (<65) patients with sporadic colorectal cancer. These findings inform prognostication and may direct future research for this high-risk population.

The aims of the present study were to investigate the association between overweight and cognitive ability and dementia, and to evaluate the usefulness of self-reported body mass index (BMI) in late life and various data sources commonly used in epidemiological studies to identify persons with dementia. Data were drawn from three population-based studies: the Swedish Adoption/Twin Study of Aging (SATSA), Aging in Women and Men: A Longitudinal Study of Gender Differences in Health Behaviour and Health among Elderly (the Gender Study), and the Finnish Lieto Study. In Study I, the agreement between self-reported and measured BMI over time was evaluated among 774 men and women, ages 40 to 88 years at baseline (mean age 63.9) participating in both the questionnaire phase and in-person testing of SATSA. Latent growth curve (LGC) modeling showed a small but significant increase between self-reported and measured BMI (0.02 kg/m2/y) over time, which would probably not affect the results if self-reported BMI were used as a continuous variable in longitudinal research. In Study II, the agreement between dementia diagnoses from various sources and dementia diagnoses set at a consensus conference was evaluated. Among the 498 elderly people ages 70 to 81 at baseline (mean age 74.5) enrolled in the Gender Study, 87 were diagnosed with dementia during an eight-year period. Review of medical records and nurse evaluations yielded the highest sensitivity (0.83 and 0.80, respectively) and a high specificity (0.98 and 0.96), indicating that these sources might be good proxies of dementia, while data extraction from the Swedish Inpatient Discharge Registry underestimated the prevalence of dementia (sensitivity 0.26). In Study III, the association between being overweight in midlife and cognitive ability in late life was examined in SATSA. The 781 participants ages 25 to 63 at baseline (mean age 41.6) in 1963 or 1973 self-reported their height and weight. From 1986 until 2002, they were assessed five times using a cognitive test battery. LGC models showed that people with higher midlife BMI scores had significantly lower cognitive ability and a significantly steeper decline than their thinner counterparts, an association that persisted when those who developed dementia during the study period were excluded from the analysis. This finding indicates that being overweight might affect cognitive ability independently of dementia. In Study IV, the association between BMI and dementia risk in older persons was described among 605 persons without dementia and ages 65 to 92 at baseline (mean age 70.8) in the Lieto Study. Among these, 86 persons were diagnosed with dementia during eight years of follow-up. Cox regression analyses indicated that for each unit increase in BMI score, the risk of dementia decreased 8% (hazard ratio = 0.92, 95% confidence interval = 0.87–0.97) and the association remained significant when individuals who developed dementia during the first four years of follow-up were excluded from the analyses. This result suggests that low BMI scores are present almost a decade before clinical dementia onset.

Stroke is a major cause of premature death and disability throughout the world and atrial fibrillation (AF) is one of the most common preventable causes of stroke. AF affects about 10% of individuals aged ≥80 years, but warfarin is substantially under-used in this age group despite being effective in preventing AF-related thromboembolic events. AF-related ischaemic strokes tend to be severe and incur high care costs, and non-cerebral systemic embolism secondary to AF is also a major clinical burden. Despite that, there are few population-based studies on AF-related ischaemic stroke, and no recent study of the burden of AF-related thromboembolism and the population impact of under-treatment. I have used data from the Oxford Vascular Study (OXVASC), a prospective, population-based incidence study of vascular disease in all territories, which was started in April 2002 and is on-going. The study population comprises of 92,728 individuals registered with 100 family physicians in nine general practices and uses multiple overlapping methods of “hot” and “cold” pursuit to achieve near-complete ascertainment of all patients with acute vascular events. There are several findings described by the research in this thesis which have important implications for public health and can be utilised to improve secondary prevention in stroke. First, I have shown that one-third of all incident embolic events were related to AF and 60% of AF-related embolic events occurred at ≥80 years. Second, I have shown that only 9% of patients aged ≥80 years with incident embolic event related to known prior AF were on premorbid warfarin, and consequently three quarters of those previously independent were dead or disabled six months post event. Third, I have shown that there has been no reduction in age-specific incidence of AF-related ischaemic stroke in Oxfordshire over the last 25 years. Fourth, I have shown that assuming age-specific incidence does not continue to rise, if prevention is not improved, the number of embolic events at age ≥80 years would be expected to treble by 2050 (72,975 AF-related embolic events), with 84% of events at all ages occurring at age ≥80. Fifth, I have shown through a meta-analysis that one in five incident strokes had a history of prior AF of which only 19% were on premorbid warfarin, and AF was related to one in three incident ischaemic strokes. Sixth, I have shown that 1 in 5 stroke patients with known prior AF subsequently became institutionalised and incurred high acute and long-term care costs. Seventh, I have shown that one in five patients with undetermined cerebral ischaemic event subsequently had AF-related late recurrent stroke. Eighth, I have shown that even though TIA or ischaemic stroke patients who subsequently turned out to have new AF at follow-up had significantly higher baseline NT-proBNP compared to non-AF group, its utility is limited by low sensitivity and specificity. Ninth, I have shown in another meta-analysis that the duration of cardiac monitoring after cerebral ischaemic events was the main determinant of the observed rate of pAF, and that 5-7 days of monitoring may be adequate in unselected patient populations. Finally, I have shown that using 5-day event loop recording in clinic patients with TIA and minor ischaemic stroke could detect 12% new AF and the delay in monitoring did not reduce the sensitivity of pAF detection.

Abstract Sciatica is a common musculoskeletal disorder, especially among the working-age population. It causes huge costs to society through work absenteeism and hospital treatments. The common cause of sciatica is the herniated lumbar disc compressing the nerve root. The neurological deficit resulting from this compression can be either sensory or motor-related. Earlier studies have shown both genetic and environmental factors to exist in the aetiology of sciatica. The aim of this study was to investigate how individual, physical and work-related risk factors associate with sciatica, and to determine the prevalence of sciatic pain among Finnish adolescents. The data used in this study were from large, population-based Finnish surveys: the Northern-Finland Birth Cohort 1986, the Mobile Clinic Health Examination Survey, the Mini-Finland Health Survey, Health 2000, the Young Finns Study, and the Helsinki Health Studies. The results of this study showed that sciatic symptoms are already common among adolescents. Young women in particular reported sciatic pain more often than men. Low back pain at the age of 16 predicted sciatic pain at the age of 18. Physically demanding work, smoking and obesity were found to predict an increased risk of hospitalization for sciatica. In the Mobile Clinic Health Examination Survey, leisure-time physical activity protected men from hospitalization for sciatica, whereas among women obesity increased this risk. Occupation also modified the effects of various risk factors. In the Mini-Finland Health Survey, overweight and obese participants who had been exposed to whole body vibration in their work were at a higher risk of hospitalization due to sciatica. In addition, lifting and carrying heavy objects at work, or sedentary work involving the handling of heavy objects predicted an increased risk of hospitalization for sciatica. Heavy or very heavy work protected against hospitalization for sciatica. A meta-analysis of four prospective cohort studies showed that walking and cycling to work reduced the risk of hospitalization for sciatica. This thesis extends our knowledge regarding the risk factors for sciatica. Its results show that these risk factors are complex and can modify each other’s effects. Thus, further research on the interactions of the various risk factors is needed. / Tiivistelmä Iskias on etenkin työikäisten keskuudessa yleinen sairaus, joka aiheuttaa paljon työkyvyttömyyttä ja sairaalahoitoja ja siten kustannuksia yhteiskunnalle. Iskiaksen taustalla on yleensä välilevyn pullistuma, joka painaa hermojuuria. Hermojuuren puristuksesta aiheutuva neurologinen puutosoire voi olla joko sensorinen tai motorinen. Aiemmat tutkimukset ovat osoittaneet iskiaksen taustalla olevan sekä geneettisiä että ympäristötekijöitä. Tämän tutkimuksen tavoitteena oli selvittää yksilöllisten, fyysisten ja työhön liittyvien riskitekijöiden yhteyttä iskiakseen sekä tutkia iskiaskivun yleisyyttä suomalaisilla nuorilla aikuisilla. Aineistoina tutkimuksessa käytettiin laajoja väestöpohjaisia suomalaisia aineistoja: Pohjois-Suomen syntymäkohorttia 1986, Autoklinikka- ja Mini-Suomi-aineistoja sekä Terveys 2000-, Young Finns- ja Helsinki Health Study -tutkimuksia. Tutkimuksessa selvisi, että iskiasoireet ovat yleisiä jo nuorilla aikuisilla ja etenkin naiset raportoivat iskiaskipua useammin miehiin verrattuna. 16-vuotiaana ilmennyt alaselkäkipu oli myös yhteydessä iskiasoireisiin 18 vuoden iässä. Työn fyysisen rasittavuuden, tupakoinnin ja lihavuuden todettiin lisäävän sairaalahoitoon johtaneen iskiaksen riskiä. Autoklinikka-aineistossa vapaa-ajan fyysinen aktiivisuus suojasi miehiä sairaalahoitoon johtaneelta iskiakselta, naisilla taas ylipaino lisäsi iskiaksen riskiä. Lisäksi ammatti näytti muokkaavan riskitekijöiden assosiaatioita. Mini-Suomi-aineistossa ylipainoiset ja lihavat henkilöt, jotka olivat altistuneet työssään koko kehon tärinälle, olivat korkeammassa riskissä sairaalahoitoon johtaneen iskiaksen suhteen. Töissä raskaiden esineiden nostaminen tai kantaminen tai painavien esineiden käsittelyä sisältävä istumatyö lisäsivät myös iskiaksen riskiä. Raskas tai erittäin raskas fyysinen työ taas näytti suojaavan sairaalahoitoon johtaneelta iskiakselta. Neljän prospektiivisen kohorttitutkimuksen meta-analyysissä selvisi, että työmatkapyöräily tai -kävely vähensivät sairaalahoitoon johtaneen iskiaksen riskiä. Tämä väitöstutkimus kasvattaa ymmärtämystä iskiaksen taustalla olevista riskitekijöistä. Tulokset osoittavat, että iskiaksen riskitekijät ovat moniulotteisia ja voivat muokata toistensa vaikutuksia, joten lisätutkimusta eri riskitekijöiden välisistä interaktioista tarvitaan.

Introdução: O crescente aumento da população idosa faz aumentar a necessidade de recursos de saúde, entre eles o uso de medicamentos.Objetivo: Estudar os riscos de polifarmácia em idosos no município de São Paulo, Brasil. Métodos: Este estudo faz parte do projeto SABE – Saúde,Bem-estar e Envelhecimento – através de questionários por amostra em domicílios de 2143 idosos com 60 anos e mais composta por sorteio. Os dados finais foram ponderados e expandidos de modo que representem a população idosa no ano de 2000. A polifarmácia foi definida como o uso de quatro ou mais medicamentos, e utilizado o estudo de regressão logística por passos (IC 95%). Resultados: A média do número de medicamentos foi de 2,72 e a prevalência de polifarmácia de 31,5%. A polifarmácia foi mais prevalente em mulheres com 75 anos e mais (52,1%), religião espírita(51,2%), que declaram estado de saúde ruim (40,2%) e escolaridade acima de 12 anos (46,9%). Verificou-se que 71,1% adquirem medicamentos do próprio bolso, 15,95% se automedicam e a não adesão é devida ao custo (9,1%). Os riscos para polifarmácia foram mulheres (OR 2,2), idade acima de 75 anos (OR 1,5), consulta e internação em quatro meses (OR de 1,9 e 3,8) e problemas cardíacos (OR 3,8). Quanto ao medicamento impróprio a prevalência foi de 15,6%. Conclusão: Os riscos identificados na polifarmácia mostram uma necessidade de políticas públicas que visem promover o uso racional de medicamentos. / Introduction: The continuous growth of the elderly population increases the need for further health resources; amongst them is the use of drugs.Object: Study the risks’ Polypharmacy in the population of elderly people within the city of São Paulo, Brasil. Method: This study is part of the ‘SABE’ project – Health, Well-being and aging. This survey is carried out by using a sample questionnaire in the residence of 2143 people aged 60 and over. The final data are pondered and expanded to represent the population of elderly people in the year 2000. In order to analyze, polypharmacy was defined as four or more drugs, using the study of stepwise logistical regression (IC95%). Results: The average number of drug stays at 2,72, with a prevalence of polyfarmacy of 31,5%. Polyfarmacy is more prevalent amongst women aged 75 and over (52,1%); spiritualists (51,2%); those who claim poor self perceived health status (40,2%); those whose level of education is at least 12 years (46,9%). It has been observed that 71,1% use their own money to buy drugs; 15,9% practice self-medication; the cost of treatment being the cause of nonadherence (9,1%). Women are more at risk (OR 2,2), aged 75 and over (OR 1,5), visit to the physician and hospitalization within four months(OR from 1,9 to 3,8), cardiovascular conditions (OR 3,8). As for inappropriate use of medications, the prevalence is 15,6%. Conclusion: The identified risks in polypharmacy show a need for public policies that would promote a more rational use of medications.

In cases of forensic dental identification, a key factor in the comparison of the dental status of deceased persons with antemortem dental records is the matching of dental restorations in individual teeth. Many studies have been performed showing the prevalence of dental interventions. This has mostly been performed by counting the numbers of decayed, missing and filled teeth (DMFT) in each mouth without detailed data collection on a per-tooth basis. The purpose of this study was to investigate the research question: to what extent would data on the distribution and prevalence of restoration types in the human dentition facilitate forensic identification? A database program was developed to allow efficient collection and collation of dental trait information. Provision was made for storing information relating to a subject's individual teeth, such as restorative materials used and surfaces filled. Other data, such as missing teeth, caries status on a per-individual tooth basis and presence and details of types of prostheses may be stored. iii Data from patients attending a private group practice in Hobart was collected and a system was devised to enable the likelihood of dental trait occurrence to be calculated in cases of forensic dental identification. The capabilities of the system are demonstrated in a series of mock cases of dental identification. An opportunity to make use of the database for which it was designed arose in relation to the analysis of person identification evidence in a murder trial at the Tasmanian Supreme Court. The use of this reference database in evidence invoked lengthy debate involving the judge, crown prosecutor and defence barristers. The resulting voire dire was resolved in favour of conclusions drawn from the use of the database being admitted in evidence. The legal precedent set in the Marlow trial may possibly offer encouragement for practising odontologists to further the concept of establishing and using reference databases of dental traits in population groups in other parts of Australia.

Background There is a widespread perception that the health status of rural Australians is poorer than that of urban Australians, characterised by higher mortality, lower life expectancies, and an increased incidence of some diseases. At present this perception is difficult to confirm or refute, in terms of mental illness, because of limited published data on the extent of mental illness in regional Australia. Australians from rural areas are also reported to have less access to appropriate health care compared to their urban counterparts; however, there is limited evidence to support such claims using large population-based epidemiological data. It is not known whether remoteness per se is an important determinant of health. Aim To determine if rural and remote South Australians were disadvantaged in terms of their mental health status and access to health care. The aims were to: 1) determine if prevalence of mental illness and comorbidity were associated with accessibility and remoteness; 2) examine the effects of accessibility and remoteness on health service utilisation; and 3) determine if remoteness per se was an important determinant of mental illness. Methods Prospectively designed, secondary analysis of data from a large cross sectional, population–based health survey, conducted in South Australia (SA) in 2000. In all, 2,454 adults, aged 18 years or more, were randomly selected and interviewed using the Computer Assisted Telephone Interview (CATI) system. CATI is a telephone monitoring system that is an efficient means of assessing self-reported aspects of population health, particularly in rural and remote areas. Psychological distress and depression were assessed using the Kessler 10 (K10) Psychological Distress Scale, the SF-12 measure of health status and self-reported, medically-confirmed mental illness, in the previous 12 months. Additional outcome measures included socio-demographic characteristics, a range of health services measures, psychosocial and health risk factors. Geographical variation in outcome measures was assessed using the Accessibility and Remoteness Index of Australia (ARIA). The data were analysed using SPPS and Stata statistical programs and weighted by region, age, sex and probability of selection in the household, using the 1999 total estimated resident population (ERP) figures supplied by the Australian Bureau of Statistics. Direct age-sex standardisation was applied to prevalence rates of mental illness, socio-demographic and health service utilisation data. Results Overall age-sex adjusted mental illness prevalence estimates were similar using the three measures of psychological distress (10.5%), depression (12.9%) and self-reported medically-confirmed mental illness (12.9%). For each measure, there was no significant variation in prevalence across ARIA categories, except for a lower than expected prevalence of depression (7.7%) in the accessible category. There was also no significant difference in the median number of uses of four types of health services across ARIA categories. Significantly fewer residents of highly accessible areas reported never using primary health care services (14.4% vs. 22.2% in very remote areas), and significantly more reported high use (6 visits, 29.3% vs. 21.5%). Fewer residents of remote areas reported never attending hospital (65.6% vs. 73.8% in highly accessible areas). Frequency of use of mental health services was low and not significantly different across ARIA categories. Very remote residents were more likely to spend at least one night in a public hospital (15.8%) than were residents of other areas (eg 5.9% for highly accessible areas). After controlling for the joint effects of stressful life events, perceived control of life events, socio-demographic characteristics and health risk factors, odds of mental illness did not vary by ARIA category (highly accessible: reference category; accessible: OR 0.9, 95% CI 0.60-1.31; moderately accessible: OR 0.80, 95% CI 0.45-1.43; remote/ very remote: 0.70, 95% CI 0.44-1.03). The most important predictors of mental illness in the multivariate logistic model were female sex; smoking; low consumption of vegetables; low exercise; a physical condition; perceived lack of control with: life in general, personal life, job security or health; and major stressful events such as family or domestic violence and the death of someone close. Conclusions Prevalence rates of psychological distress, depression and medically-confirmed mental illness in SA were high. However, there was no evidence that the prevalence of these conditions varied substantially across ARIA categories. The frequency of use of a range of health services was also broadly similar across the state. Remoteness per se was not associated with mental illness, either directly or indirectly as an important confounder in stressful life event/mental illness associations. Psychosocial factors were more important determinants of mental illness. The data do not support existing stereotypes of a rural – urban mental health differential in SA and point to potential mechanisms that may be responsible for poorer mental health outcomes.

<p>Olfactory function is impaired in aging and dementia. The general aim of this thesis was to identify variables that predict olfactory function and dysfunction (assessed with an odor identification test) in middle-aged and elderly adults. The thesis investigated whether odor identification performance was associated with demographic variables, cognitive function, the ApoE gene, dementia, and other health-related variables. The ApoE-ε4 allele is associated with cognitive impairment and Alzheimer’s disease, the most common form of dementia. The studies included in this thesis used data from the Betula study, a large-scale, population-based prospective study on aging, memory, and health. Study 1 investigated demographic and cognitive predictors of odor identification ability in non-demented participants. The results showed that younger age, female sex, and high education contributed to better odor identification ability. Cognitive speed and vocabulary had a small additional influence. Study 2 included information about ApoE genotypes, dementia and other health-related variables. The results indicated that the ApoE-ε4 allele was associated with odor identification impairment among the elderly, but not middle-aged adults. Participants who were demented at the time of testing or became demented within five years after testing exhibited olfactory impairments. Interestingly, the age-related olfactory impairment in ε4-carriers was independent of clinical dementia within five years. In Study 3, decline in global cognitive status over a five-year test-retest interval was predicted in a sample of elderly participants. The major result was a three-way interaction reflecting that odor identification impairment, old age, in combination with the ε4 allele predicted a larger cognitive decline. However, odor identification impairment did not predict cognitive change in elderly who were non-carriers of the ε4 allele. Overall, the results indicate that odor identification impairment in elderly is related to ApoE-ε4, cognitive decline, and clinical and pre-clinical stages of dementia. Theoretical and practical implications of the results are discussed. Furthermore, it is proposed that in order to effectively predict clinical dementia or cognitive decline from olfactory assessment in the elderly, variables that mediate (e.g. neuropathology) or moderate (e.g. age) the associations between olfactory function, the ε4 allele, and dementia need to be further evaluated, preferably in studies using longitudinal assessment.</p>

<p>The aim was to contribute to the optimal use of the resting ECG by exploring, in middle-aged and elderly men, the development and regression of ECG abnormalities; the prognostic value of the ECG for cardiovascular disease compared to conventional risk factors; and the impact of age at baseline and follow-up time for prediction of cardiovascular disease.</p><p>It was based on the Uppsala Study of Adult Men cohort that was started in 1970. Participants were examined at ages 50, 70, 77, and 82, with annual updates on mortality and in-hospital morbidity using national registries. </p><p>The studies indicated that the prevalence of silent MI and frequency of regression of major Q/QS patterns may be higher than previously believed. Considering that persistent T wave abnormalities and ST segment depression carried twice as high a risk for future cardiovascular disease (CVD) mortality as new or reverted abnormalities, the results suggested that serial electrocardiograms (ECG) would contribute to proper risk assessment. Also, the inclusion of ischemic ECG findings significantly increased the predictive power of the Framingham score at age 70 for CVD. </p><p>While hypertension and dyslipidemia were consistent long-term risk factors for myocardial infarction at ages 50 and 70, the length of follow-up period and age at baseline affected the predictive power of ECG abnormalities, fasting insulin, BMI, and smoking. </p><p>For stroke, midlife values for blood pressure and ECG abnormalities retained prognostic value over long follow-up periods, even though they improved when re-measured in elderly participants. ApoB/apoA1 ratio, driven by apoA1, was associated with stroke in elderly but not middle-aged men. Hyperinsulinemia and diabetes mellitus were more specifically associated with ischemic stroke than with any-cause stroke. </p><p>In summary, the resting ECG carried prognostic information beyond conventional risk factors. Even though the low prevalence of ECG abnormalities at the age of 50 calls into question the role of the ECG as a screening tool, the additional risk information it carries with it justifies its regular and repeated registration above the age of 50. </p>

Olfactory function is impaired in aging and dementia. The general aim of this thesis was to identify variables that predict olfactory function and dysfunction (assessed with an odor identification test) in middle-aged and elderly adults. The thesis investigated whether odor identification performance was associated with demographic variables, cognitive function, the ApoE gene, dementia, and other health-related variables. The ApoE-ε4 allele is associated with cognitive impairment and Alzheimer’s disease, the most common form of dementia. The studies included in this thesis used data from the Betula study, a large-scale, population-based prospective study on aging, memory, and health. Study 1 investigated demographic and cognitive predictors of odor identification ability in non-demented participants. The results showed that younger age, female sex, and high education contributed to better odor identification ability. Cognitive speed and vocabulary had a small additional influence. Study 2 included information about ApoE genotypes, dementia and other health-related variables. The results indicated that the ApoE-ε4 allele was associated with odor identification impairment among the elderly, but not middle-aged adults. Participants who were demented at the time of testing or became demented within five years after testing exhibited olfactory impairments. Interestingly, the age-related olfactory impairment in ε4-carriers was independent of clinical dementia within five years. In Study 3, decline in global cognitive status over a five-year test-retest interval was predicted in a sample of elderly participants. The major result was a three-way interaction reflecting that odor identification impairment, old age, in combination with the ε4 allele predicted a larger cognitive decline. However, odor identification impairment did not predict cognitive change in elderly who were non-carriers of the ε4 allele. Overall, the results indicate that odor identification impairment in elderly is related to ApoE-ε4, cognitive decline, and clinical and pre-clinical stages of dementia. Theoretical and practical implications of the results are discussed. Furthermore, it is proposed that in order to effectively predict clinical dementia or cognitive decline from olfactory assessment in the elderly, variables that mediate (e.g. neuropathology) or moderate (e.g. age) the associations between olfactory function, the ε4 allele, and dementia need to be further evaluated, preferably in studies using longitudinal assessment.