loading page

Lecture Notes on Epidemiology
  • Arindam Basu
Arindam Basu
University of Canterbury
Author Profile

Abstract

Epidemiology is defined as the study of the distribution and determinants of health related states and the use of that knowledge to prevent diseases and promote wellness. Environmental epidemiology is a branch of Epidemiology where we study the Environment as a determinant of human health.  Environment is a cause of our health and wellness. By "cause" we mean that environmental factors as events or phenomenon precede our health states and are associated with it. Our health states and illnesses are caused by a number of factors and there is no one on one association with a specific environmental correlate and health state. Because of this, we should think of health related states and causes of health related states where causes are components of a larger model and the model with its component causes is the sufficient unit to cause disease. How do determine that an exposure E is a cause of a health outcome O? In order to decide if E is a cause of O, we need to first establish that E has a valid association with O. By valid association, we mean that we should explain that the association we observe between E and O cannot be explained by chance alone, we have to rule out that possibility. We also have to eliminate any possibility of biased observation or reporting that can result in an apparent association between E and O but in reality that association is spurious. We will need to  control for any potential or actual confounding variable, that is, a variable that is associated both with E and O, but does not come in any causal pathway connecting the two. After we have established a valid association, we will then evaluate that the nature of this association is one of cause and effect. A way to test that association is to use Hill's considerations (often referred to as "Hill's criteria"). 
Cause and effect or non-causal effects for that matter, need to be studied using measurements of diseases or health related states and measures of associations. Measures of health related states include incidence rates and prevalence. Incidence and prevalences must be standardised for age or other variables when more than one population or the same population is studied over time. Measures of association include rate ratios or relative risk estimates, odds ratios, and in case of Environmental health, we also use impact fractions or attributable risks to indicate the role of an environmental factor on a health outcome in the population.  
We obtain these measures by conducting one or more epidemiological studies. Environmental epidemiological study designs are observational study designs and fall in the categories of ecological studies, cross-sectional surveys, case control studies and cohort studies. A range of analytical strategies are used to analyse data from epidemiological studies. These include simple linear models and linear regression, logistic regression for case control and cross-sectional surveys, survival analysis and Cox Proportional Hazards model for Cohort Studies. 
Epidemiological studies form the basis of understanding Environmental health and therefore forms a good groundwork for understanding the processes of environmental health impact assessment. Environmental investigations are also based on epidemiological principles. For thinking about all environmental health related issues, it is a good idea to start with the facts, develop an explanatory theory and identify hyotheses based on these theories and collect facts to validate the theories by testing hypotheses.