PUBH6005 Epidemiology Assessment 2: Study Design, Sampling and Risk

This report analyzes epidemiological study designs, sampling techniques, and population risk factors, drawing on the Whitehall studies and other relevant research. Part 1 examines the Whitehall I and II studies, focusing on their sampling frames, methodologies (including cohort studies, statistical analyses like chi-square tests and logistic regression), and generalizability of findings regarding socioeconomic status, work stress, and health outcomes. It discusses how similar studies could be conducted in the Australian context, referencing the 45 and Up Study cohort. Part 2 explores various observational research designs (cohort, cross-sectional, and longitudinal studies) and experimental designs (randomized controlled trials), assessing their advantages, disadvantages, and ethical dilemmas. It provides examples of how each design could be applied to specific research questions, such as the association between binge eating and depression, the long-term effects of detention, and the impacts of folate supplementation on pregnancy and autism development. The report emphasizes the importance of understanding causal associations, potential biases, and ethical considerations in epidemiological research.

Running head: EPIDEMIOLOGY
Assessment 2: Study design, sampling and population risk
Name of the Student
Name of the University
Author Note

1EPIDEMIOLOGY
Part 1
1. The sampling frame used for the first Whitehall I study involved male British civil
servants in order to determine the risk factors for cardiorespiratory disease. This
initial prospective cohort study was conducted in order to examine more than 17,500
men civil servants who were aged between 20 and 64, and the research was conducted
over a timespan of ten years, commencing from 1967. This study was based on
drawing a comparison regarding mortality of individuals who were employed in the
civil services and also demonstrated that the rate of mortality was higher amid low
grade servants, in comparison to high grade servants. In contrast, the second cohort
study was the Whitehall II study was conducted between 1985 and 1988, with the aim
of exploring the health status of 10,308 civil servants, who were aged 35-55 years. Of
them one third were females and two third were males (Marmot et al., 1991). This
study investigated the association between stress, work and health of the servants.
2. Breeze et al. (2001) used four different measures for determining morbidity risks of
the participants namely, general poor health, poor physical performance, poor mental
health, and disability. The researchers also used the Short Form 36 Health Survey
(SF-36) to measure the health status of the patients, in addition to classifying the
participants based on their employment grade. Statistical analysis was conducted by
chi-square tests that helped in determining univariate association. The researchers also
performed Logistic regression (Stata5 for Windows 3.1) for estimating the odds ratio
and confidence interval for each outcome. Data collection by Chandola et al. (2008)
encompassed measuring the self-reported work stress with the use of the job-strain
questionnaire. The researchers used cox proportional hazard regression models for
exploring the association between CHD event incident and work stress measures,
following adjustment for different variables. They also used linear/logistic regression

2EPIDEMIOLOGY
model for determining the association between the continuous CHD risk factors and
work stress. Furthermore, the MONICA criteria was used for defining non-fatal MI.
In contrast, Marmot et al. (1978) used multivariate analysis using multiple logistic
equation for estimating the impact of each risk factor. Records were obtained from
Central Registry of the National Health Service for exploring the rate of mortality
during a particular period. They also calculated the comparative increase in risk.
3. The results of all the studies can be generalised to the wider population, owing to the
fact that social gradient, and its impact on physical and mental health outcomes is not
a phenomenon that is predominantly confined to a particular nation. Breeze et al.
(2001) stated that the socioeconomic status of a person in middle age, and at about
retirement age is related with an increase in morbidity in old age. This finding can be
generalised to the wider population since socioeconomic status creates an impact on
the general human functioning, such as the mental and physical health (Harrell, 2018).
Findings from Chandola et al. (2008) suggested that work stress acts in the form of a
significant determinant for coronary heart disease amid people who are of the working
age, which in turn gets mediated through impacts created on stress pathways and
health behaviours. This can be generalised since work related stress can contribute to
hypertension, cardiovascular complications, diabetes, and obesity (Siegrist,
Wahrendorf & Siegrist, 2016). Marmot et al. (1978) suggested that an increase in
CHD mortality is reported by men who belong to the working class. However, work
stress has been more associated with female gender (Nelson & Burke, 2018).
4. It would be feasible to conduct a similar study in the Australian context, by recruiting
participants belonging to the 45 and up study cohort owing to the fact that it
comprises of people who are aged 45 years or more, and had been created to explore
reliable evidences on several outcomes and exposure that were of public health

3EPIDEMIOLOGY
significance. This cohort had been conducted by recruiting 250,000 males and
females aged 45 years and more from the overall population of New South Wales
State in Australia. The major objective of conducting a similar study would be to
determine the impacts of socioeconomic gradient and stress on health outcomes of the
residents (Mihrshahi et al., 2017). Hence, the cohort can be used since it focuses on
several priority areas that are namely, health impacts, impacts of environmental
factors, economic and social determinants, risk factors, health in aged people, and
usage of healthcare services.
Part 2
1. Causal association will be determined using observational research design, since it
will help in drawing inferences from a population sample where there is no control of
the researcher on the independent variable, owing to logistical or ethical concerns.
Some advantages of this design are that it enables assessing research participants in
natural setting, thus providing insights about the causal association between
dependent and independent variables that are under investigation (VanderWeele &
Ding, 2017). In addition, the researchers will also be able to modify vantage point, in
relation to real-time variables. However, the researchers will have little or no control
on the environment and there remains a high possibility for subjective bias.
Confidentiality and autonomy of participants are major ethical dilemma.
2. A cohort study will be conducted to determine association between binge eating and
depression and obese individuals. This kind of study will be conducted in order to
determine how particular risk factors increase the likelihood of people to suffer from
adverse health outcomes. It will allow to calculate incidence and effects of single or
multiple exposure (Kooijman et al., 2016). However, the subjects need to be followed
for a longer time, and the research can lead to introduction of bias. One major ethical

4EPIDEMIOLOGY
dilemma will be related to dual relationship between the participants and the
researchers.
3. In order to determine the impact of long term effects of detention, a cross-sectional
study will be conducted among the asylum seekers. The primary advantage of this
research is that it will help in routine collection of data, which in turn will help in
proving a particular hypothesis. One major disadvantage of this design is that will not
provide a description of which particular variable is the effect, and which is the cause
(Creswell & Poth, 2017). One major ethical concern will be seeking informed consent
from the participants.
4. Longitudinal study needs to be conducted for determining impacts of folate
supplementation on pregnancy and autism development. Unlike cross-sectional study,
a longitudinal study will track differences in same individuals over a prolonged
period. The advantages are that it will help in exploring viable impacts over time, will
ensure validity and focus, and will provide high accuracy at the time of observing the
changes (Creswell & Creswell, 2017). Nonetheless, it would consume huge time, and
would also increase risks for panel attrition. Confidentiality and informed consent can
be two major ethical dilemmas.
5. Randomised controlled trial needs to be conducted to determine the impact of the
drug. The design will help in making causal inferences and will also minimise
selection bias and allocation bias. Furthermore, blinding will also minimise
performance bias, besides making the findings more interpretable. However, power
calculation might demand huge samples, and long trial might lead to absence of
relevance of the results (Creswell & Creswell, 2017). Ethical dilemma might
encompass autonomy at the time of allocation concealment or blinding both the

5EPIDEMIOLOGY
researchers, as well as the participants to the intervention that they will be subjected
to.
References
Breeze, E., Fletcher, A. E., Leon, D. A., Marmot, M. G., Clarke, R. J., & Shipley, M. J.
(2001). Do socioeconomic disadvantages persist into old age? Self-reported morbidity
in a 29-year follow-up of the Whitehall Study. American journal of public
health, 91(2), 277.
Chandola, T., Britton, A., Brunner, E., Hemingway, H., Malik, M., Kumari, M., ... &
Marmot, M. (2008). Work stress and coronary heart disease: what are the
mechanisms?. European heart journal, 29(5), 640-648.
Creswell, J. W., & Creswell, J. D. (2017). Research design: Qualitative, quantitative, and
mixed methods approaches. Sage publications.
Creswell, J. W., & Poth, C. N. (2017). Qualitative inquiry and research design: Choosing
among five approaches. Sage publications.
Harrell, C. (2018). Socioeconomic Status and Health: The Protective Role of Religiosity
among African Americans.
Kooijman, M. N., Kruithof, C. J., van Duijn, C. M., Duijts, L., Franco, O. H., van IJzendoorn,
M. H., ... & Moll, H. A. (2016). The Generation R Study: design and cohort update
2017. European journal of epidemiology, 31(12), 1243-1264.
Marmot, M. G., Rose, G., Shipley, M., & Hamilton, P. J. (1978). Employment grade and
coronary heart disease in British civil servants. Journal of Epidemiology &
Community Health, 32(4), 244-249.
Marmot, M. G., Stansfeld, S., Patel, C., North, F., Head, J., White, I., ... & Smith, G. D.
(1991). Health inequalities among British civil servants: the Whitehall II study. The
Lancet, 337(8754), 1387-1393.

6EPIDEMIOLOGY
Mihrshahi, S., Ding, D., Gale, J., Allman-Farinelli, M., Banks, E., & Bauman, A. E. (2017).
Vegetarian diet and all-cause mortality: Evidence from a large population-based
Australian cohort-the 45 and Up Study. Preventive medicine, 97, 1-7.
Nelson, D. L., & Burke, R. J. (2018). Gender, Work Stress, and Health.
Siegrist, J., Wahrendorf, M., & Siegrist. (2016). Work stress and health in a globalized
economy. Springer Verlag.
VanderWeele, T. J., & Ding, P. (2017). Sensitivity analysis in observational research:
introducing the E-value. Annals of internal medicine.