Risk of prevalent and incident dementia associated with insulin-like growth factor and insulin-like growth factor-binding protein 3.

Insulin-like growth factor 1 (IGF-1) influences cell proliferation and survival. In the extracellular environment, IGF-1 circulates bound to proteins (IGF-binding proteins; IGFBP), some of which have physiological effects that seem independent of IGF-1, including the brain (for example, IGFBP-3). We completed a systematic review of the association between dementia and IGF-1 and IGFBP-3, and a cross-sectional and longitudinal study designed to investigate if lower plasma concentration of these proteins increased the risk of prevalent and incident dementia. A total of 3967 men aged 71-89 years joined the study, of whom 535 (13.5%) showed evidence of prevalent cognitive impairment. The plasma concentrations of IGF-1 and IGFBP-3 were similar for men with and without cognitive impairment. The 3432 men free of cognitive impairment were then followed for up to 13 years. During this time 571 (16.6%) developed dementia. The plasma concentration of IGF-1 had no association with incident dementia. The doubling of the plasma concentration of IGFBP-3 decreased the hazard ratio of dementia by 23% (95% confidence interval=5-37%). The results were not affected by age, body mass index and history of smoking, diabetes, hypertension, coronary heart disease or stroke. If these findings are confirmed by others, the plasma concentration of IGFBP-3 could be used to improve the accuracy of predictive models of dementia and as a potential new factor to assist in the development of prevention and treatment strategies.

Profile of diabetes in men aged 79-97 years: the Western Australian Health in Men Study.

AIMS: To investigate behavioural, physical and biochemical characteristics associated with diabetes in the oldest age group of elderly men. METHODS: We conducted a cross-sectional analysis of community-dwelling men aged 79-97 years from Perth, Western Australia. Lifestyle behaviours, self-rated health, physical function, and fasting glucose and HbA1c levels were assessed. RESULTS: Of 1426 men, 315 had diabetes (22%). Men with diabetes were of similar age to men without (84.9 vs 84.5 years; P = 0.14). Only 26.5% of men with diabetes self-rated their health as excellent or very good, compared with 40.6% of men without diabetes (P < 0.001). Diabetes was associated with less involvement with recreational walking (32.7 vs 41.0%; P < 0.01) and leisure activities (19.0 vs 26.5%; P < 0.01). Men with diabetes had poorer physical function on multiple measures, including longer times for the Timed Up-and-Go test (15.0 ± 6.9 s vs 13.4 ± 5.3 s; P < 0.001) and weaker knee extension (20.2 vs 21.9 kg; P < 0.001). In multivariate analyses, diabetes was associated with an increased prevalence of myocardial infarction (odds ratio 1.80, 95% CI 1.25-2.60; P < 0.001) and falls resulting in injury (odds ratio 1.55, 95% CI 1.06-2.26; P = 0.02). Average HbA1c was 49 ± 8 mmol/mol (6.6 ± 0.8%) in men with diabetes, with 90.6% of these men on diet or oral hypoglycaemic therapy. CONCLUSIONS: In older men, diabetes is associated with poorer self-perceived health, reduced healthy lifestyle behaviours and physical function, heart disease and injurious falls. The majority of these men with diabetes had good glycaemic control. Encouraging healthy lifestyle behaviours and improving physical function should be evaluated as interventions to improve quality-of-life and health outcomes.

Cerebral small vessel disease, medial temporal lobe atrophy and cognitive status in patients with ischaemic stroke and transient ischaemic attack.

BACKGROUND AND PURPOSE: Small vessel disease (SVD) and Alzheimer's disease (AD) are two common causes of cognitive impairment and dementia, traditionally considered as distinct processes. The relationship between radiological features suggestive of AD and SVD was explored, and the association of each of these features with cognitive status at 1 year was investigated in patients with stroke or transient ischaemic attack. METHODS: Anonymized data were accessed from the Virtual International Stroke Trials Archive (VISTA). Medial temporal lobe atrophy (MTA; a marker of AD) and markers of SVD were rated using validated ordinal visual scales. Cognitive status was evaluated with the Mini Mental State Examination (MMSE) 1 year after the index stroke. Logistic regression models were used to investigate independent associations between (i) baseline SVD features and MTA and (ii) all baseline neuroimaging features and cognitive status 1 year post-stroke. RESULTS: In all, 234 patients were included, mean (±SD) age 65.7 ± 13.1 years, 145 (62%) male. Moderate to severe MTA was present in 104 (44%) patients. SVD features were independently associated with MTA (P < 0.001). After adjusting for age, sex, disability after stroke, hypertension and diabetes mellitus, MTA was the only radiological feature independently associated with cognitive impairment, defined using thresholds of MMSE ≤ 26 (odds ratio 1.94; 95% confidence interval 1.28-2.94) and MMSE ≤ 23 (odds ratio 2.31; 95% confidence interval 1.48-3.62). CONCLUSION: In patients with ischaemic cerebrovascular disease, SVD features are associated with MTA, which is a common finding in stroke survivors. SVD and AD type neurodegeneration coexist, but the AD marker MTA, rather than SVD markers, is associated with post-stroke cognitive impairment.

Determinants of post-stroke cognitive impairment: analysis from VISTA.

BACKGROUND: Post-stroke cognitive impairment (PSCI) occurs commonly and is linked with development of dementia. We investigated the relationship between demographic, clinical and stroke symptoms at stroke onset and the presence of PSCI at 1 and 3 years after stroke. METHODS: We accessed anonymized data from the Virtual International Stroke Trial Archive (VISTA), including demographic and clinical variables. Post-stroke cognitive impairment was defined as a Mini-Mental State Examination (MMSE) score of ≤26. We assessed univariate relationships between baseline stroke symptoms and PSCI at 1 and 3 years following stroke, retaining the significant and relevant clinical factors as covariates in a final adjusted logistic regression model. RESULTS: We analysed data on 5435 patients with recent (median 33 days) stroke or transient ischaemic attack (TIA). Mean (±SD) age was 62.6 (±12.6) years; 3476 (65%) patients were male. Follow-up data were available for 2270 and 1294 patients at 1 and 3 years, respectively. At 1 year, 781 (34%) patients had MMSE≤26; at 3 years, 391 (30%) had MMSE≤26. After adjusting for age, stroke severity, hypertension, diabetes and type of qualifying event, initial stroke impairment (leg paralysis) was associated with increased rate of PSCI at 1 year (OR=1.62; 95% CI=1.20-2.20) and at 3 years (OR=1.95; 95% CI=1.23-3.09). Associations were consistent on subgroup analysis restricted to ischaemic stroke and transient ischaemic attack (N=4992). CONCLUSIONS: Besides well-known determinants of PSCI such as age, stroke severity and the presence of vascular risk factors, also leg paralysis is associated with subsequent of PSCI up to 3 years after stroke.

The triangular association of ADH1B genetic polymorphism, alcohol consumption and the risk of depression in older men.

Alcohol use, particularly alcohol abuse and dependence, are associated with increased risk of depression. Current diagnostic criteria suggest that the relationship is causal, but the evidence has only been derived from observational studies that are subject to confounding and bias. Given the logistic and ethical constraints that would be associated with a trial of alcohol use to prevent depression, we aimed to complete a Mendelian randomization study to determine if a genetic polymorphism associated with alcohol abuse and dependence (ADH1B rs1229984 G-->A) contributed to modulate the risk of depression in a community-derived cohort of older men. This retrospective analysis of a cohort of 3873 community-dwelling men aged 65-83 years living in the metropolitan region of Perth, Western Australia, investigated the triangular association between the rs1229984 G-->A polymorphism and alcohol use and, after 3.2-8.2 years, the presence of current depression or history of depression. The mean number of standard drinks consumed per week (n; standard deviation; range) according to genotype was AA 1.8 (17; 2.7; 0-7), GA 5.9 (262; 7.5; 0-35), GG 8.5 (3594; 10.9; 0-140) (GG>AA, GG>GA; P<0.001). Consumption of 1 or 2 drinks per day decreased the odds of depression (n=610) by 30 and 40%, whereas consumption of more than six drinks daily more than doubled the odds of depression (odds ratio: 2.12, 95% confidence interval: 1.02, 4.40). The ADH1B rs1229984 G-->A polymorphism was not associated with current or past depression (P=0.857). In addition, the presence of the A allele did not interact with the alcohol use to modulate the risk of depression (P=0.725). These results suggest that alcohol consumption does not cause or prevent depression in older men.

Stroke prevention with rivaroxaban in higher-risk populations with atrial fibrillation.

BACKGROUND: Atrial fibrillation (AF), the most common cardiac arrhythmia, is a major risk factor for stroke. Rivaroxaban, an oral factor Xa inhibitor, is approved for the prevention of stroke in patients with non-valvular AF. In the pivotal phase III trial ROCKET AF, rivaroxaban demonstrated non-inferiority compared with warfarin for reducing the risk of stroke or systemic embolism (SE) in patients with AF (intention-to-treat analysis), without an increased risk of major bleeding. Superior efficacy vs. warfarin was achieved while patients were on study medication. Other direct oral factor Xa inhibitors have completed phase III clinical trials in this indication. Compared with warfarin, apixaban (in the ARISTOTLE trial) and edoxaban (in the ENGAGE-AF trial) were shown to be superior or non-inferior, respectively, for reduction in stroke or SE risk in patients with AF. Baseline stroke risk, as indicated by CHADS2 scores, was lower in patients in the ARISTOTLE and ENGAGE-AF trials than in ROCKET AF. OBJECTIVES: This review discusses the main findings from ROCKET AF, specifically examining recent subgroup analyses investigating rivaroxaban use across various patient types at high risk for adverse outcomes, including those with prior stroke or transient ischaemic attack, reduced renal function, prior myocardial infarction, peripheral artery disease, heart failure or patients aged ≥ 75 years and those resident in East Asia. CONCLUSIONS: These subgroup analyses demonstrate that the treatment effect for rivaroxaban vs. warfarin is broadly consistent across a wide range of patient groups, with respect to both efficacy and safety.

Economic evaluation of the Very Early Rehabilitation in SpEech (VERSE) intervention.

INTRODUCTION: There is limited evidence on the costs and outcomes of patients with aphasia after stroke. The aim of this study was to estimate costs in patients with aphasia after stroke according to the aphasia therapies provided. METHODS: A three-arm, prospective, randomized, parallel group, open-label, blinded endpoint assessment trial conducted in Australia and New Zealand. Usual ward-based care (Usual Care) was compared to additional usual ward-based therapy (Usual Care Plus) and a prescribed and structured aphasia therapy program in addition to Usual Care (the VERSE intervention). Information about healthcare utilization and productivity were collected to estimate costs in Australian dollars for 2017-18. Multivariable regression models with bootstrapping were used to estimate differences in costs and outcomes (clinically meaningful change in aphasia severity measured by the WAB-R-AQ). RESULTS: Overall, 202/246 (82%) participants completed follow-up at 26 weeks. Median costs per person were $23,322 (Q1 5,367, Q3 52,669, n = 63) for Usual Care, $26,923 (Q1 7,303, Q3 76,174, n = 70) for Usual Care Plus and $31,143 (Q1 7,001. Q3 62,390, n = 69) for VERSE. No differences in costs and outcomes were detected between groups. Usual Care Plus was inferior (i.e. more costly and less effective) in 64% of iterations, and in 18% was less costly and less effective compared to Usual Care. VERSE was inferior in 65% of samples and less costly and less effective in 12% compared to Usual Care. CONCLUSION: There was limited evidence that additional intensively delivered aphasia therapy within the context of usual acute care provided was worthwhile in terms of costs for the outcomes gained.

Peripheral arterial disease increases the risk of subsequent hip fracture in older men: the Health in Men Study.

UNLABELLED: The aim of the present study was to assess whether peripheral arterial disease is associated with an increased risk of hip fracture in a cohort of 12,094 older men. There was no association between claudication and hip fracture, but there was a significant association with an ankle brachial index (ABI) <0.9. INTRODUCTION: It is uncertain whether peripheral arterial disease (PAD) is associated with an increased risk of subsequent hip fracture. The aim of the present study was to assess this in a large cohort of men aged 65 years and over. METHODS: Claudication was assessed by means of the Edinburgh Claudication Questionnaire in 12,094 men, and the ABI was measured in 4,321 of these men. Hospitalisations with hip fracture were identified by record linkage. The association between both claudication and an ABI <0.9 and subsequent hip fractures was assessed using survival curves and Cox regression models. RESULTS: Amongst the 12,094 men, the baseline prevalence of claudication according to the ECQ was 5.3 %. Amongst the 4,321 men with ABI results, the prevalence of an ABI <0.9 was 11.7 %. Of the 506 men with an ABI <0.9, 129 (25.5 %) also had claudication. Over a median (range) follow-up of 10.8 (0.3-12.7) years, 343 (2.8 %) of the 12,094 men were admitted to hospital with a hip fracture. There was no association between claudication and subsequent hip fractures (hazard ratio (HR) = 0.95; 95 % confidence interval (CI), 0.60, 1.52). Over a median (range) follow-up of 11.1 (0.06-12.3) years 135 (3.1 %) of the 4,321 men with ABI data were admitted to hospital with hip fractures. There was a significant association between an ABI <0.9 and subsequent hip fracture (HR = 1.69; 95 % CI, 1.08, 2.63). CONCLUSION: Older men with PAD defined as ABI < 0.9 are at increased risk of hip fracture, whereas the symptom of claudication is not an independent predictor of hip fracture.

Homocysteine, methylenetetrahydrofolate reductase C677T polymorphism and cognitive impairment: the health in men study.

High total plasma homocysteine (tHcy) has been associated with cognitive impairment in later life, but it is unclear if this association is causal or is due to confounding. The C677T polymorphism of the 5,10 methylenetetrahydrofolate reductase gene (MTHFR) increases basal tHcy, but its contribution to cognitive impairment has not been established. We designed this study to determine if tHcy is causally related to cognitive impairment in later life by investigating its association with high tHcy and the MTHFR-C677T polymorphism. We recruited 1778 older men from the Health in Men Study cohort and established caseness on the basis of the participants' scores on a Telephone Interview for Cognitive Status score 27 in 2008. Exposure to tHcy, gene status and other variables of interest were obtained from assessments 4-7 years earlier. Multivariate logistic regression showed that the odds of cognitive impairment increased with a doubling of tHcy (adjusted odds ratio, OR 1.36; 95% confidence interval, 95% CI 1.02-1.82). Compared with the wild CC genotype, participants with the MTHFR-TT genotype had 46% greater odds of cognitive impairment (OR 1.46, 95% CI 1.01-2.11, P=0.043). The results of this study are consistent with, but do not prove the hypothesis that high tHcy causes cognitive impairment in later life.

Is hypovitaminosis D associated with abdominal aortic aneurysm, and is there a dose-response relationship?

OBJECTIVE: This study aims to investigate the association between plasma 25-hydroxyvitamin D (25(OH)D) concentrations with the presence of abdominal aortic aneurysm (AAA) and aortic diameter. DESIGN: An observational study of 4233 community-dwelling men aged 70-88 years, who participated in a randomised controlled trial of screening for AAA. METHODS: Infrarenal aortic diameter measured by ultrasound and 25(OH)D by immunoassay. RESULTS: A total of 311 men (7.4%) with AAA (defined as aortic diameter ≥ 30 mm) comprised the study. Multivariable models were adjusted for age, smoking, cardiovascular disease, hypertension, diabetes, dyslipidaemia, body mass index and serum creatinine concentration. Amongst men with the lowest 25(OH)D quartile of values compared with the highest quartile, the adjusted odds ratio of having an AAA increased in a graded fashion from 1.23 (95% confidence interval (CI) 0.87-1.73) for AAA ≥ 30 mm to 5.42 (95% CI 1.85-15.88) for AAA ≥ 40 mm. Similarly, there was a dose-response relationship between 25(OH)D concentrations and the size of the AAA: every 10-nmol l(-1) decrease in 25(OH)D levels was associated with 0.49 mm (95% CI 0.11-0.87) increase in mean aortic diameter. CONCLUSIONS: Low vitamin D status is associated with the presence of larger AAA in older men, and there is a graded inverse relationship between 25(OH)D concentrations and AAA diameter. Further research is needed to clarify the mechanisms underlying these associations.

Hearing Impairment and Incident Frailty in Later Life: The Health in Men Study (HIMS).

OBJECTIVES: This study is designed to determine if hearing loss is associated with increased risk of frailty in later life. DESIGN: A prospective cohort study. SETTING AND PARTICIPANTS: We retrieved data of a community sample of men aged 70 years and above living in the metropolitan region of Perth, Western Australia. 3,285 participants who were free of frailty at the beginning of the study were followed for up to 17 years. Data were retrieved from the Health in Men Study (HIMS) and the Western Australian Data Linkage System (WADLS). MEASUREMENTS: Hearing loss was defined by self-report or by diagnosis recorded in the WADLS. Incident frailty was assessed using the Hospital Frailty Risk Score (HFRS). RESULTS: A total of 2,348 (71.5%) men developed frailty during follow up. The adjusted hazard ratio was 1.03 (95% CI: 0.95-1.12). The majority of the participants became frail by age 90 regardless of hearing condition. The time point where half of the group become frail was delayed by 14.4 months for men without hearing loss compared with hearing impaired men. CONCLUSIONS: Hearing loss is not associated with incident frailty in men aged 70 years or older when frailty was measured by HFRS. However, this statistically non-significant result could be due to the low sensitivity of study measures. Also, we found a trend that men with hearing loss were more likely to develop frailty compared with their normal-hearing peers, suggesting a potential association between hearing loss and frailty.

Associations of total osteocalcin with all-cause and cardiovascular mortality in older men. The Health In Men Study.

SUMMARY: In older men, both lower and higher total osteocalcin levels predict increased all-cause mortality, with comparable associations for cardiovascular and non-cardiovascular deaths. Differences in osteocalcin levels might influence glucose metabolism and thereby cardiovascular risk, or reflect changes in bone turnover thus representing a marker for poorer health outcomes. INTRODUCTION: Reduced levels of total osteocalcin (TOC) are associated with adiposity, insulin resistance and type 2 diabetes, implying this bone-derived peptide might modulate cardiovascular risk. However, there are few longitudinal data relating TOC levels to survival. We examined associations of TOC level with all-cause and cardiovascular mortality in older men. METHODS: We conducted a prospective cohort study of community-dwelling men aged 70-89 years. Aliquots of plasma collected at baseline (2001-2004) were assayed for TOC. Incidence and causes of death to 31 December 2008 were ascertained using data linkage. Cox regression analyses were performed with adjustment for conventional cardiovascular risk factors. RESULTS: From 3,542 men followed for median 5.2 years there were 572 deaths (16.1%). Mortality was lowest in men with TOC levels in the second quintile (12.6%). In multivariate analyses, men with TOC in the lowest and highest quintiles of values had increased all-cause mortality (Q1 vs Q2: hazard ratio [HR], 1.36; 95% confidence interval 1.02-1.80 and Q5 vs Q2: HR, 1.53, 95% CI 1.18-1.98). Men with low TOC levels had similar HR for cardiovascular and non-cardiovascular deaths (Q1 vs Q2: HR, 1.35 and 1.30 respectively). Higher TOC levels predicted cardiovascular disease (CVD)-related mortality (Q5 vs Q2, HR, 1.69, 95% CI 1.09-2.64). CONCLUSIONS: TOC predicts all-cause and CVD-related mortality in community-dwelling older men. However, the relationship is U shaped with men at both ends of the distribution at increased risk. Further investigation is required to clarify whether the underlying mechanisms involve altered bone turnover or relate specifically to the biological activity of osteocalcin.

Assessing the risk of stroke from neck manipulation: a systematic review.

BACKGROUND: Strokes, typically involving vertebral artery dissection, can follow cervical spinal manipulative therapy, and these types of stroke occur rarely. There is disagreement about whether a strong association between neck manipulation and stroke exists. An earlier systematic review found two relevant studies of association that used controls, which also discussed the limitations of the two papers. Our systematic review updates the earlier review, and aims to determine whether conclusive evidence of a strong association exists. METHODS: PRISMA guidelines for systematic reviews were followed, and the literature was searched using a strategy that included the terms 'neck manipulation' and 'stroke' from the PubMed, Embase, CINAHL Plus and AMED databases. Citations were included if they met criteria such as being case-control studies, and dealt with neck manipulation and/or neck movement/positioning. Papers were scored for their quality, using similar criteria to the earlier review. For individual criteria, each study was assigned a full positive score if the criterion was satisfied completely. RESULTS: Four case-control studies and one case-control study, which included a case- crossover design, met the selection criteria, but all of them had at least three items in the quality assessment that failed to be completely positive. Two studies were assessed to be the most robustly designed, one indicating a strong association between stroke and various intensities of neck movement, including manipulation, and the other suggesting a much reduced relative association when using primary care practitioners' visits as controls. However, potential biases and confounders render the results inconclusive. CONCLUSION: Conclusive evidence is lacking for a strong association between neck manipulation and stroke, but is also absent for no association. Future studies of association will need to minimise potential biases and confounders, and ideally have sufficient numbers of cases to allow subgroup analysis for different types of neck manipulation and neck movement.

Bayesian methods for meta-analysis of causal relationships estimated using genetic instrumental variables.

Genetic markers can be used as instrumental variables, in an analogous way to randomization in a clinical trial, to estimate the causal relationship between a phenotype and an outcome variable. Our purpose is to extend the existing methods for such Mendelian randomization studies to the context of multiple genetic markers measured in multiple studies, based on the analysis of individual participant data. First, for a single genetic marker in one study, we show that the usual ratio of coefficients approach can be reformulated as a regression with heterogeneous error in the explanatory variable. This can be implemented using a Bayesian approach, which is next extended to include multiple genetic markers. We then propose a hierarchical model for undertaking a meta-analysis of multiple studies, in which it is not necessary that the same genetic markers are measured in each study. This provides an overall estimate of the causal relationship between the phenotype and the outcome, and an assessment of its heterogeneity across studies. As an example, we estimate the causal relationship of blood concentrations of C-reactive protein on fibrinogen levels using data from 11 studies. These methods provide a flexible framework for efficient estimation of causal relationships derived from multiple studies. Issues discussed include weak instrument bias, analysis of binary outcome data such as disease risk, missing genetic data, and the use of haplotypes.

When the patient fails to respond to treatment: TIAs that go on, and on.

Most transient ischaemic attacks (TIAs) stop, with or without treatment, but some are followed by stroke within days or longer. But if the TIAs do not stop then the diagnosis must be reviewed (are they really TIAs or could they be migraine or epilepsy?), and if they are TIAs, what are they caused by (atherothromboembolism, embolism from the heart, etc)? With this information, both the medical and any surgical treatment can be optimised, even though one must accept that the randomised controlled trials mostly have not addressed the particular issue of continuing TIAs.

Beyond Stroke Prevention in Atrial Fibrillation: Exploring Further Unmet Needs with Rivaroxaban.

With improved life expectancy and the aging population, the global burden of atrial fibrillation (AF) continues to increase, and with AF comes an estimated fivefold increased risk of ischaemic stroke. Prophylactic anticoagulant therapy is more effective in reducing the risk of ischaemic stroke in AF patients than acetylsalicylic acid or dual-antiplatelet therapy combining ASA with clopidogrel. Non-vitamin K antagonist oral anticoagulants are the standard of care for stroke prevention in patients with non-valvular AF. The optimal anticoagulant strategy to prevent thromboembolism in AF patients who are undergoing percutaneous coronary intervention and stenting, those who have undergone successful transcatheter aortic valve replacement and those with embolic stroke of undetermined source are areas of ongoing research. This article provides an update on three randomized controlled trials of rivaroxaban, a direct, oral factor Xa inhibitor, that are complete or are ongoing, in these unmet areas of stroke prevention: oPen-label, randomized, controlled, multicentre study explorIng twO treatmeNt stratEgiEs of Rivaroxaban and a dose-adjusted oral vitamin K antagonist treatment strategy in patients with Atrial Fibrillation who undergo Percutaneous Coronary Intervention (PIONEER AF-PCI) trial; the New Approach riVaroxaban Inhibition of factor Xa in a Global trial vs Aspirin to prevenT Embolism in Embolic Stroke of Undetermined Source (NAVIGATE ESUS) trial and the Global study comparing a rivAroxaban-based antithrombotic strategy to an antipLatelet-based strategy after transcatheter aortIc vaLve rEplacement to Optimize clinical outcomes (GALILEO) trial. The data from these studies are anticipated to help address continuing challenges for a range of patients at risk of stroke.

Depression as a modifiable factor to decrease the risk of dementia.

Depression is an accepted risk factor for dementia, but it is unclear if this relationship is causal. This study investigated whether dementia associated with depression decreases with antidepressant use and is independent of the time between exposure to depression and the onset of dementia. We completed a 14-year longitudinal study of 4922 cognitively healthy men aged 71-89 years, and collected information about history of past depression, current depression and severity of depressive symptoms. Other measures included use of antidepressants, age, education, smoking and history of diabetes, hypertension, coronary heart disease, and stroke. The onset of dementia and death during follow-up was ascertained via the Western Australian Data Linkage System. A total of 682 men had past (n=388) or current (n=294) depression. During 8.9 years follow-up, 903 (18.3%) developed dementia and 1884 (38.3%) died free of dementia. The sub-hazard ratios (SHRs) of dementia for men with past and current depression were 1.3 (95% confidence interval (CI)=1.0, 1.6) and 1.5 (95% CI=1.2, 2.0). The use of antidepressants did not decrease this risk. Compared to men with no symptoms, the SHRs of dementia associated with questionable, mild-to-moderate and severe depressive symptoms were 1.2 (95% CI=1.0, 1.4), 1.7 (95% CI=1.4, 2.2) and 2.1 (95% CI=1.4, 3.2), respectively. The association between depression and dementia was only apparent during the initial 5 years of follow-up. Older men with history of depression are at increased risk of developing dementia, but depression is more likely to be a marker of incipient dementia than a truly modifiable risk factor.