1.5 Principles of epidemiology and the relevance of toxicological data Flashcards Preview

B01. Principles of toxicology and epidemiology > 1.5 Principles of epidemiology and the relevance of toxicological data > Flashcards

Flashcards in 1.5 Principles of epidemiology and the relevance of toxicological data Deck (25)
Loading flashcards...

Toxicology is

the science of adverse effects of chemical substances on living organisms.


Toxicological studies aim to assess the adverse effects related to different doses in
order to find this ‘acceptably safe’ level.

The work is carried out in two phases:

1) By collecting data on the properties of chemicals, results of studies and accidental misuse of chemicals.
2) By predicting the effects of chemicals in different situations.


Human toxicity information should be generated whenever possible by means other
than vertebrate animal tests, for example: 3

 in vitro methods
 qualitative or quantitative structure-activity relationship models
 information from structurally related substances (grouping or read-across).


The use of acute toxicity data is mainly to label and classify chemicals based on their
toxicity, for application to the human situation, however as mentioned there have
been numerous challenges to the approach, on a range of grounds including: 5

 Human exposures are more likely to be repetitive low-doses than a single massive dose as per animal tests.

 Different species deal with and react to chemicals differently including the rates and routes of metabolism and in absorption, distribution, and excretion; the target organs involved; and sensitivity to toxicity.

 There is often a high degree of variability in acute toxicity data obtained from rodents of different ages, sexes, and genetic strains. Environmentally dependent variables such as: weight, diet, temperature and humidity also influence results.

 Animal testing is costly and time-consuming, and can delay the timely regulation of human health protection.

 Animal tests for acute toxicity have never been validated to modern standards.


The NOAEL is

the highest point on the exposure-response curve at which no
adverse health effects are observed.


The LOAEL is

the lowest point on the exposure-response curve at which adverse
health effects are observed.


Ames test description

The test uses a strain of Salmonella typhimurium that carries a defective (mutant) gene making it unable to synthesize the amino acid histidine (His) from the ingredients in its culture medium.

The altered Salmonella strains are combined in a test tube with the chemical of interest and animal liver enzymes which detect what might happen if the chemical entered a human body.

The Salmonella are then transferred to a petri dish to grow for one or two days. The altered Salmonella used for the test require the amino acid histidine to grow, and a positive result in the test indicates that the test substance has induced a back
mutation in the Salmonella meaning it no longer requires histidine to grow.


QSAR models are

Quantitative Structure Activity Relationships (QSAR) are mathematical, computer based, models which are designed to predict the physico-chemical properties, human health and environmental effects of a substance from knowledge of its chemical structure (molecular descriptors).


‘Read across’ is

a technique of filling data gaps. To ‘read across’ is to apply data from a tested chemical for a particular property or effect (cancer, reproductive
toxicity, etc.) to a similar untested chemical.


Epidemiology is

the study of the distribution and determinants of health-related states or events in specified populations, and the application of this study to
the control of health problems.


Epidemiology has an important role notably
in: 3

 Establishing the causes and determinants of occupational ill-health.

 Ensuring adequate recognition and quantification of this.

 Determining appropriate occupational exposure limits.


Cross sectional studies

In a cross-sectional study, the prevalence of a particular disease or of a set of
symptoms or other indication of ill-health is investigated in a single time-point (or
over a relatively narrow period of time). Comparisons can then be made in the
frequency of ill-health for example between workers exposed to a particular hazard,
and those who are not, or between workers experiencing different degrees of

A cross sectional study can determine the prevalence rate, which is defined as the
number of existing cases of disease divided by the population at a specified time

For example if a chest X-ray survey of quarry workers is conducted it might show
that workers in quarries with high exposure to quartz (a crystalline form of silica)
might have a higher prevalence of pneumoconiosis than those in quarries with little
or no such exposure.


Cohort studies

The study begins with a group of people free of disease (or outcome of concern). The group is sub-classified according to exposure to a suspected cause (for example: smokers and non-smokers) and followed over time to determine the development of new cases of disease (for example: lung cancer) in each group.

The incidence rate (number of new cases divided by population exposed) of disease in both groups can be calculated and the relative risk (RR) of disease in the exposed group compared to the non-exposed group can be determined.

A relative risk of 3 – 4 (i.e. the exposed group is 3 to 4 times more likely to develop the disease of concern, than the non-exposed group) is considered to be a good indication of a causal relationship between cause and effect, although on its own is no guarantee of a causal relationship.


Cohort studies may be conducted prospectively which means

they start in the present and track exposures and outcomes into the future. This poses ethical problems as if a cause and effect relationship is suspected an immediate intervention is the appropriate course of action. This would protect individuals but would ruin the natural experiment so that no meaningful data would be generated


Cohort studies are often conducted retrospectively

The study is conducted in exactly the same way as a prospective study but the starting point is some time in
the past.


case control study

starts with a population of individuals with the disease under investigation (Cases) and compares them with a carefully matched control group of
individuals without the disease (Controls).

A food poisoning outbreak at a party provides a practical example of a case-control study. Of a hundred guests 60 developed food poisoning. These are the cases and the 40 who did not get food poisoning are the controls. The range of foods at the buffet are the suspected causal agents and by interviewing each guest (cases and controls) the odds of each group having consumed each food type can be calculated and the Odds Ratio might show that those who suffered food poisoning
(cases) were four times more likely to have ate the chicken drumsticks, for example, than those who did not get food poisoning (controls).


Inferring causation (3 steps)

Demonstrate a statistical relationship between
cause and effect (RelativeRisk or Odds Ratio)

Eliminate non-casual
● Confounding
● Bias
● Chance

Apply causal criteria to substantiate casual relationship


non-causal explanations (list)






results from multiple associations between the exposure, the disease and some third factor (the
'confounding variable') which is associated with both the exposure and independently affects the risk of developing the disease.

An example of confounding is the observed association between air pollution and cardiac or pulmonary disease. There now appears to be little doubt that a causal association exists between say particulate air pollution and respiratory morbidity and mortality. However, an unsophisticated study simply relating air pollution to ill health and death might lead to the conclusion that the association is much stronger than it really is.

This is because of confounding variables such as temperature.

Low temperatures in winter may contribute to increased mortality. In addition, low temperatures (in meteorological conditions of inversion) may favour increased pollution levels. If the confounding caused by temperature is taken into account i.e.
it is resolved, then the association between air pollution and health becomes weaker.


Selection and participation bias: This occurs if

the study populations being compared are not strictly comparable, for example: in a study to determine the effect of a Workplace Health Promotion (WHP) programme on 'sickness absence', the rate of subsequent sickness absence might have been compared between
those who participated in the WHP programme and those who did not.

What if the results appeared to show that the WHP group had lower rates of sickness absence? Would one be entitled to conclude that WHP resulted in less sickness absence?


Observation bias: This occurs if

non-comparable information is obtained from
each study group, for example: if one was conducting a case-control study to determine whether scleroderma (systemic sclerosis) was associated with occupational exposure to certain hazards (such as trichloroethylene or silica),
bias could arise if the interviewer knew (or could tell) which people were the
cases and which were the controls. He/she might seek more detail about exposures from the cases than the controls who did not have the disease. Thus, observer bias would have influenced the results.


Recall bias might arise if

the cases (suffering from the disease) having
previously pondered about possible causes of their misfortune, were to recollect
more detail about their past exposures, than the controls (who may have no real
motivation to reflect at length on their past occupations).



Imagine that the frequency of back pain among employees in a particular workplace needed to be determined. Rather than questioning all the employees, it would be easier to administer questionnaires to only a sample of this population and from them, estimate the frequency of back pain in the workers. As a
consequence it must be borne in mind that chance may affected the results because of random variation in the population. It could be that, by chance, the selected sample was a particularly fit and healthy group and consequently the frequency of back pain in the workplace population at large may be underestimated.


Criteria for determining causation

Sir Austin Bradford-Hill established nine key criteria that should be considered before concluding that a relationship is causal as long ago as 1965. These have been reviewed and revised extensively over the last fifty years yet remain
essentially the same. (list all 9)



Strength of association


Exposure and dose


Biologic plausibility




The fundamental problem with epidemiological studies is that

they cannot prove a causal relationship. A high relative risk (or odds ratio) can strongly infer causation but this is not the same as saying that exposure to agent X causes disease Y. The
scientific evidence shows a strong link between smoking and lung cancer, however cigarette smoking is neither a necessary cause (people who have never smoked get lung cancer) nor sufficient cause (people who smoke all their lives might not
get lung cancer) of lung cancer