A cross-sectional study design to conduct a multi-level exposure control experiment

(d) SLOT Case Control Intake Yes 54 203 Exposure No 106 736 Crude Odd ratio for domestic violence= [(54)/(106)]/[(203)/(736)] = 1.85 Therefore, the crude odds ratio is calculated to be 1.85 (e) The adjusted value of odd ration in the table was found to be 1.44. (f) According to table 5, the effect of physical violence on low birth weight was only 13 individuals out of

Running Head: INTRODUCTION TO EPIDEMIOLOGY
INTRODUCTION TO EPIDEMIOLOGY
Name of the Student
Name of the University
Author Note

Running Head: INTRODUCTION TO EPIDEMIOLOGY
Table of Contents
Answer 1:.........................................................................................................................................3
Answer 2:.........................................................................................................................................6
Answer 3:.........................................................................................................................................8
Answer 4:.......................................................................................................................................10
References:....................................................................................................................................12

Running Head: INTRODUCTION TO EPIDEMIOLOGY
Answer 1:
(a)
This paper follows a cross-sectional study design to conduct its study. This study design
measures the exposure and results of the participants studied at the same time (Vo et al., 2019).
The sample size in this study was calculated by following the sampsi command inside
STATA 13.0 (Software for Statistics and Data Sciences). Data collection tools used WHO
questionnaire for domestic violence which is useful in a cross-sectional study. Statistical tables
did the calculation of the data with significance and confidence intervals. The above-stated
factors prove that the methodological characteristics support the above-stated study design.
(b)
The primary exposures of the study include interviewing the participants and checking
the exposure to domestic violence. Standard deviation and Mean was calculated from the
affected population. The study was divided into four parts- Domestic, Emotional, Sexual and
Physical violence. Percentage and Frequencies of the participants affected by all the violence
were tabulated. The percentages of the relation of domestic violence with the age group and
educational background of women were calculated.
The outcomes of this study involve a statistical idea about the exposure of women to
domestic violence. This data is further related to the association of domestic violence with lower
birth weights.

Running Head: INTRODUCTION TO EPIDEMIOLOGY
(c)
The risk of selection bias is low in this study. Selection bias stands for the selection of
individuals for a study where proper randomisation during data collection and result calculation
is not achieved (Munafò et al., 2017). Ten districts were selected at random out of 24 districts,
and the study was performed. Age group also covered a vast range (18 to above 40 years).
(d)
SLOT Case Control
Exposur
e Yes
54 203
Exposur
e
No
106 736
Crude Odd ratio for domestic violence= [(54)/(106)]/[(203)/(736)] = 1.85
Therefore, the crude odds ratio is calculated to be 1.85
(e)
The adjusted value of odd ration in the table was found to be 1.44. The calculated value
was 1.85. This difference occurred due to other variables along with domestic violence, that was
taken for the study. According to table 4, other variables that were included are the age group of
wives, occupation, housing status, economic status and husbands personality. These variables are
called confounder variables ().

Running Head: INTRODUCTION TO EPIDEMIOLOGY

Running Head: INTRODUCTION TO EPIDEMIOLOGY
(f)
According to table 5, the effect of physical violence on low birth weight was less since
only 13 individuals out of 160 test individuals reported to be associated with physical violence.
The adjusted prevalence ratio was found to be 0.69. This value is comparatively less than 1.44
for domestic violence. The reason for this less value is that, physical violence was not the only
factor in reducing its prevalence ratio. Sexual and environmental violence also influenced the
value. Therefore the value is less.
(g)
As stated in the paper, measurement bias arose from a cross-sectional study and usage of
the questionnaire to interview events that occurred previously. Authors tried to minimise the
effect by using close questions stated by WHO’s DV and used well-trained professionals as
interviewers.

Running Head: INTRODUCTION TO EPIDEMIOLOGY
Answer 2:
(a)
Categories Number of mothers
associated with the child with
a birth defect
Number of the mothers
related with healthy child
Intake of Folic acid(YES) 17 3
Intake of Folic acid(NO) 55 70
Crude Odd Ratio= [ 17
3 ]
[ 55
70 ] =7.26
Therefore the strength of the association between folic acid intake and birth defects is very high.
(b)
The name of the measure is the calculation of the crude odd ratio. Higher the value is,
more likely; the two situations are associated with each other (Ahlbom, 2017). A value higher
than one states that the intake of folic acid is directly associated with birth defects. According to
this value (7.26), it can be stated that there is a very high association of intake of folic acid with
the presence of congenital disabilities in the child.

Running Head: INTRODUCTION TO EPIDEMIOLOGY
(c)
According to the table drawn in part (a) , the total number of mothers of child with birth
defects is 72.
Out of 72, 17 of them were found to be associated to folic acid intake.
The proportion of people for whom the folic acid intake can be prevented is= ( 17
72 )∗100
Therefore the proportion = 23.61

Running Head: INTRODUCTION TO EPIDEMIOLOGY
Answer 3:
(a)
Cases of D Persons-years/1000 Risk ratio
Total
Exposed 51 3116
[ 51
3116
345
31787 ] =1.6
Not exposed 345 31787
Young adults <65
Exposed 19 1168
[ 19
1168
177
13177 ]=1.23
Not exposed 177 13,177
Old adults
Exposed 33 1948
[ 33
1948
167
18560 ] =1.77
Not exposed 167 18560
Table: Incidence of the disease by exposure status, stratified by age group (young /old)

Running Head: INTRODUCTION TO EPIDEMIOLOGY

Running Head: INTRODUCTION TO EPIDEMIOLOGY
For the above table, the association of exposure to the disease is 1.6. However, when
young adults were taken, the crude association rate decreases to 1.23, whereas for old adults, it
increases to 1.77. Therefore these two variables can be stated as confounders or effect measure
modifiers which affects the original value, 1.6.
(b)
Since the adjusted risk ratio was found to be 1.51 and the calculated risk ration equal to
1.6, the age variable can be assessed as a confounder which deviates from the original value. The
value of this variable influences both the rare exposure and the disease is resulting in an
association. This association can never be calculated with correlation and confounding is
described to be a casual concept.
(c)
Attributable risk is defined as the difference between the probability of the disease in
people exposed to it and the probability of disease in the unexposed ones (Hoang et al., 2015).
From the above table, the following risk ratios can be calculated:
The risk ratio for exposed older people= 3 3
1948
Therefore Risk ratio for exposed older people=0.017
The risk ratio for unexposed older people= 167
18560
Therefore Risk ratio for theunexposed older people=0.009
Attributable Risk ratio= ( 0.017−0.009 )∗100
= 0.8%

Running Head: INTRODUCTION TO EPIDEMIOLOGY
Therefore, the attributable risk of exposure for older people was found to be 0.8%.
(d)
From the result obtained from 3(c), it can be seen that the attributable risk of exposure for
older adults is 0.8%. According to this table, the risk of exposures for older adults is higher than
that for young adult. This condition proved that old age people were more exposed to the disease
than the younger adults. However, the risk of exposure is not high for older adults.
Answer 4:
(a)
In epidemiology, biases mean a systematic fault that finally results in the wrong
estimation of an accurate value of a specific exposure on an outcome (Bhopal, 2016). There are
two main types of biases, Information bias and Selection bias (). According to the statement
provided “In a survey of the prevalence of overweight and obese participants were measure once
for their weight and once for their height in the same lab”, the bias type can be stated to be an
information bias. The bias is reported as informational since both height and weight are not taken
at the same time. The two variables were used differently in calculating the prevalence of obesity
and overweight among people in the lab. Therefore, there can be an error since both height and
weight were collected separately. Thus, the height of a person can merge with the weight of
another person during tabulation.

Running Head: INTRODUCTION TO EPIDEMIOLOGY
(b)
According to the second statement “In another survey to estimate the prevalence of
overweight and obesity participants were asked to report on their weight and height from
their head", bias here is again an information bias. This bias is said to be information bias
because the data were collected from verbal communication with the heads participating in
the survey. By chance, they can provide the wrong height and weights of themselves and thus
resulting in information bias. This biasness can be stated as a recall bias as people have to
correctly recall what they heard in the past and accurately report the values for the study
(Bland, 2015). Therefore there is always a chance for forgetting the original values and
incorporation of new values.
(c)
The above-stated problem in the case of women underestimating their weights and men
reporting their weights accurately is due to differential misclassification. This condition
occurs when the exposure becomes unequal between the candidates used in the study
(Weuve, Sagiv, & Fox, 2018). The problem also arises when the misclassification of the case
is not equal in the case of exposed and unexposed subjects. Here, women can be stated to be
more concerned about having less weight, and thus, they underestimate the weight and states
than their original weights (Antunovic, & Hardin, 2015). However, men provide accurate
weights and thus the differential misclassification occurs in this case.

Running Head: INTRODUCTION TO EPIDEMIOLOGY
References:
Ahlbom, A. (2017). Biostatistics for epidemiologists. CRC Press.
Ananth, C. V., & Schisterman, E. F. (2017). Confounding, causality, and confusion: the role of
intermediate variables in interpreting observational studies in obstetrics. American
journal of obstetrics and gynecology, 217(2), 167-175.
Antunovic, D., & Hardin, M. (2015). Women and the blogosphere: Exploring feminist
approaches to sport. International Review for the Sociology of Sport, 50(6), 661-677.
Bhopal, R. S. (2016). Concepts of epidemiology: integrating the ideas, theories, principles, and
methods of epidemiology. Oxford University Press.
Bland, M. (2015). An introduction to medical statistics. Oxford University Press (UK).
Hoang, J. K., Reiman, R. E., Nguyen, G. B., Januzis, N., Chin, B. B., Lowry, C., & Yoshizumi,
T. T. (2015). Lifetime attributable risk of cancer from radiation exposure during
parathyroid imaging: comparison of 4D CT and parathyroid scintigraphy. American
Journal of Roentgenology, 204(5), W579-W585.
Munafò, M. R., Tilling, K., Taylor, A. E., Evans, D. M., & Davey Smith, G. (2017). Collider
scope: when selection bias can substantially influence observed associations.
International journal of epidemiology, 47(1), 226-235.
Porta, M. (Ed.). (2014). A dictionary of epidemiology. Oxford university press.