The association between physical activity and risk of neurodegenerative diseases is not well established. We therefore aimed to quantify this association using meta-analytical techniques.

We searched Medline, the Cochrane Database of Systematic Reviews and Web of Science databases from 1990 to 2007 for prospective epidemiological studies of physical activity and incident dementia, Alzheimer's and Parkinson's disease. We excluded studies of physical activity and cognitive decline without diagnosis of a neurodegenerative disease. Information on study design, participant characteristics, measurement of exposure and outcome variables, adjustment for potential confounding, and estimates of associations was abstracted independently by the two investigators.

We included 16 prospective studies in the overall analysis, which incorporated 163797 non-demented participants at baseline with 3219 cases at follow-up. We calculated pooled relative risk (RR) using a random effects model. The RR of dementia in the highest physical activity category compared with the lowest was 0.72 [95% confidence interval (CI) 0.60–0.86, p<0.001], for Alzheimer's, 0.55 (95% CI 0.36–0.84, p=0.006), and for Parkinson's 0.82 (95% CI 0.57–1.18, p=0.28).

Our results suggest that physical activity is inversely associated with risk of dementia. Future studies should examine the optimal dose of physical activity to induce protection, which presently remains unclear.

The diversity of research methods applied to psychiatric disorders results in a confusing plethora of causal claims. To help make sense of these claims, the interventionist model (IM) of causality has several attractive features. First, it connects causation with the practical interests of psychiatry, defining causation in terms of ‘what would happen under interventions’, a question of key interest to those of us whose interest is ultimately in intervening to prevent and treat illness. Second, it distinguishes between predictive-correlative and true causal relationships, an essential issue cutting across many areas in psychiatric research. Third, the IM is non-reductive and agnostic to issues of mind–body problem. Fourth, the IM model cleanly separates issues of causation from questions about the underlying mechanism. Clarifying causal influences can usefully structure the search for underlying mechanisms. Fifth, it provides a sorely needed conceptual rigor to multi-level modeling, thereby avoiding a return to uncritical holistic approaches that ‘everything is relevant’ to psychiatric illness. Sixth, the IM provides a clear way to judge both the generality and depth of explanations. In conclusion, the IM can provide a single, clear empirical framework for the evaluation of all causal claims of relevance to psychiatry and presents psychiatry with a method of avoiding the sterile metaphysical arguments about mind and brain which have preoccupied our field but yielded little of practical benefit.

The role played by anxiety in the history of psychiatric epidemiology has not been well recognized. Such lack of understanding retarded the incremental growth of psychiatric research in general populations. It seems useful to look back on this history while deliberations are being carried out about how anxiety will be presented in DSM-V.

Drawing on the literature and our own research, we examined work that was carried out during and after the Second World War by a Research Branch of the United States War Department, by the Stirling County Study, and by the Midtown Manhattan Study. The differential influences of Meyerian psychobiology and Freudian psychoanalysis are noted.

The instruments developed in the early epidemiologic endeavors used questions about nervousness, palpitations, sweating, trembling, shortness of breath, upset stomach, etc. These symptoms are important features of what the clinical literature called ‘manifest’, ‘free-floating’ or ‘chronic anxiety’. A useful descriptive name is ‘autonomic anxiety’.

Although not focusing on specific circumstances as in Panic and Phobic disorders, a non-specific form of autonomic anxiety is a common, disabling and usually chronic disorder that received empirical verification in studies of several community populations. It is suggested that two types of general anxiety may need to be recognized, one dominated by excessive worry and feelings of stress, as in the current DSM-IV definition of Generalized Anxiety Disorder (GAD), and another emphasizing frequent unexplainable autonomic fearfulness, as in the early epidemiologic studies.

Studies of cognitive behavioural therapy delivered by computer (cCBT) show clinical efficacy for treating anxiety and depression, but have not focused on barriers to uptake. Potential barriers include adverse consequences, accessibility and acceptability.

An integrated systematic review was conducted of quantitative and qualitative studies and surveys from multiple electronic databases where computers delivered cCBT for anxiety or depression.

Substantial numbers of potential participants are lost prior to trials commencing with little explanation. Among trial participants, drop-outs may be higher in the cCBT groups (odds ratio 2.03, 95% confidence interval 0.81–5.09). Only a median of 56% completed a full course of cCBT and personal circumstance was a more common cause of drop-out than difficulties with the technology or social background. Risk was rarely assessed in the majority of programs. Significant staff time was needed to support clients. Therapists were more negative about cCBT than clients.

While cCBT is likely to be an effective and acceptable intervention for some people, there are barriers to its uptake that will substantially limit its impact if not addressed. These included investigating the outcome and attitudes of those who do not make it as far as cCBT trials and why so few finish a full course of cCBT.