Search

The concept of causation in epidemiology can be illuminated by situating the discussion within a more general concept of causation in biology: "a causal relationship is one that has a mechanism that by its operation makes a difference". Mechanism and difference-making are complementary, and discovery can proceed from either direction; each type of evidence can be qualitative or quantitative. An explanation becomes fully convincing only when supported by both. In biology, causation is typically stochastic and/or multiple. Multiple causation can be analysed statistically/epidemiologically, even though it is not truly (ontologically) stochastic. This requires some degree of regularity in the outcome variable, plus sufficient variation in the exposure(s). The analysis then demonstrates co-variations between exposure(s) and outcome that regularly occur. Rose's important distinction of "causes of incidence" and "causes of cases" should be reconceptualised in terms of epidemiological visibility, raising the possibility of epidemiological "dark matter"., (Copyright © 2011 Elsevier Inc. All rights reserved.)

Methods of diagrammatic modelling have been greatly developed in the past two decades. Outside the context of infectious diseases, systematic use of diagrams in epidemiology has been mainly confined to the analysis of a single link: that between a disease outcome and its proximal determinant(s). Transmitted causes ("causes of causes") tend not to be systematically analysed. The infectious disease epidemiology modelling tradition models the human population in its environment, typically with the exposure-health relationship and the determinants of exposure being considered at individual and group/ecological levels, respectively. Some properties of the resulting systems are quite general, and are seen in unrelated contexts such as biochemical pathways. Confining analysis to a single link misses the opportunity to discover such properties. The structure of a causal diagram is derived from knowledge about how the world works, as well as from statistical evidence. A single diagram can be used to characterise a whole research area, not just a single analysis - although this depends on the degree of consistency of the causal relationships between different populations - and can therefore be used to integrate multiple datasets. Additional advantages of system-wide models include: the use of instrumental variables - now emerging as an important technique in epidemiology in the context of mendelian randomisation, but under-used in the exploitation of "natural experiments"; the explicit use of change models, which have advantages with respect to inferring causation; and in the detection and elucidation of feedback. [ABSTRACT FROM AUTHOR]

Abstract Recently, many studies have found a decrease in semen quality which has increased the focus on male reproductive health. Occupational hazards are by far the best documented in reproductive epidemiological research. Generally, occupational exposures have been divided into physical exposures (heat and radiation), chemical exposures (solvents and pesticides), psychological exposures (distress), exposure to metals and welding. The recent and/or most important epidemiological studies exploring the effect of occupational exposures on semen quality and fecundity, the ability to conceive, are reviewed. The evidence for an adverse effect on male reproduction of several occupational and environmental exposures and toxicants, such as heat, ionizing radiation, inorganic lead, dibromochloropropane, ethylene dibromide, some ethylene glycol ethers, carbon disulfide and welding operations, is strongly supported in well-designed epidemiological studies. For other agents, the association is only suspected or suggested and needs further evaluation before conclusions can be drawn. It is also important to bear in mind that many workers in the non-Western world still are exposed to substances that are banned in the Western world, sometimes in high concentrations. [ABSTRACT FROM PUBLISHER]

In the epidemiologic study of reproductive capacity, the assessment of fecundity as a functional measure is complementary to approaches that focus on biomedical mechanisms and/or that use biomarkers such as semen quality. More research is needed on time trends, spatial patterns, and particular groups, especially those exposed to potentially toxic agents. Although specific research projects will always be important, much could be gained by general population surveillance, which could be introduced into existing multipurpose surveys and repeated periodically. The core measurement would be time to pregnancy, which can be carried out using a short, acceptable questionnaire that has good validity at the group level. This should be accompanied by questions on time periods of unprotected intercourse that do not end with conception, to avoid bias resulting from exclusion of relatively infertile couples. Information is also required on contraceptive failures, recent contraceptive use, and other covariates and possibly on behavioral variables, such as the degree of planning and persistence in trying to conceive, and couples’ knowledge of fertile days of the menstrual cycle. Existing statistical methods can deal with possible biases due to “accidental” pregnancies and the effects of fertility treatment. Further methodological work is needed to avoid or measure more subtle biases, for example, to determine the best way to deal with pregnancies occurring to couples whose approach to family formation is relaxed, for whom the concept of “pregnancy planning” does not apply. [ABSTRACT FROM PUBLISHER]

Health impact assessment (HIA) aims to make the health consequences of decisions explicit. Decision‐makers need to know that the conclusions of HIA are robust. Quantified estimates of potential health impacts may be more influential but there are a number of concerns. First, not everything that can be quantified is important. Second, not everything that is being quantified at present should be, if this cannot be done robustly. Finally, not everything that is important can be quantified: rigorous qualitative HIA will still be needed for a thorough assessment. This paper presents the first published attempt to provide practical guidance on what is required to perform robust, quantitative HIA. Initial steps include profiling the affected populations, obtaining evidence for postulated impacts, and determining how differences in subgroups' exposures and susceptibilities affect impacts. Using epidemiological evidence for HIA is different from carrying out a new study. Key steps in quantifying impacts are mapping the causal pathway, selecting appropriate outcome measures and selecting or developing a statistical model. Evidence from different sources is needed. For many health impacts, evidence of an effect may be scarce and estimates of the size and nature of the relationship may be inadequate. Assumptions and uncertainties must therefore be explicit. Modelled data can sometimes be tested against empirical data but sensitivity analyses are crucial. When scientific problems occur, discontinuing the study is not an option, as HIA is usually intended to inform real decisions. Both qualitative and quantitative elements of HIA must be performed robustly to be of value. [ABSTRACT FROM PUBLISHER]