8 November 2023

Analytic studies
Saad Ibrahim Al-Ghabban*
*Department of Family and Community Medicine, College of Medicine, University of Al-
Ameed, Karbala, Iraq

Analytic epidemiology encompasses the collection of data to explain observed phenomena or to

test hypotheses. Analytical epidemiology studies are the Cohort study and Case-control study; they
require information to:
a. know where to look
b. know what to control for
c. develop viable hypotheses(2)
It may be based on data from clinical studies, laboratory studies, field studies, or studies of
records.

Objectives of analytic studies:

1. Analytic studies attempt to establish a causal link between a risk factor and an outcome
(i.e. To explain observed phenomena);
2. They are designed specifically to test hypotheses that have usually been generated from
descriptive studies
panel study or follow up study
1 Cohort (or prospective) study or incidence study
Cohort is a group of people who share a common characteristic or experience e.g., People born on
a same day, Students who joined college in a year.
For cohort studies, a group of persons who are exposed to the suspected aetiological agent are
compared with matched control subjects who have not been similarly exposed. During the follow-
up period, the incidence rate of the disease among exposed is compared with the rate among non-
exposed.
For example, in a study on cancer of the lung, a large group of persons were questioned about
their smoking habits. During the follow-up period, the incidence rate of cancer of the lung among
the smokers (exposed) was compared with the rate among non-smokers (non-exposed). In this
cohort study, the subjects were selected on the basis of exposure or non-exposure.
Characteristics of cohort study:
1 Subjects should not have outcome variable on entry
2 Subjects are defined on the basis of exposure status depends on the risk factors
3 No new subjects allowed in after initial recruitment
4 Subjects are followed over time to assess disease development
5 The rates of disease incidence among the exposed and unexposed groups are determined
and compared
6 The unit of observation & analysis is the individual
7 Designed to discover whether or not there is a relationship between a particular
characteristic and a particular condition; The characteristic is suspected of playing some
part in causing the condition.

Cohort study design

A cohort or prospective study is designed to discover whether there is a relationship between a
particular characteristic and a particular condition; generally, the characteristic is suspected of
playing some part in causing the condition. The study population for a cohort or prospective study
must consist of people, who do not have the condition to be studied, but some of whom have the
characteristic and some of whom do not.

Page 1 of 7
8 November 2023

At the 1st observation point: subjects are all disease free, and exposure is used to classify subjects
into exposed or unexposed. The subjects are followed to document incidence (2nd observation
point).
It may be necessary at the outset to perform a screening or cross-sectional study to select those
who are free of the condition; they become the reference population for the study.
A sample, taken from the reference population, and then followed for a specified period of time
(which may be several years) to determine the proportion of each of the sample and control
groups who develop the condition under study. The design of cohort study is shown in Figure 1.

Figure 1 Cohort or prospective study

e.g., Framingham heart study:

Framingham study is a population-based, observational cohort study that was initiated by the
United States public health service in 1948 to prospectively investigate the epidemiology and risk
factors for cardiovascular disease. The Framingham heart study participants, and their children
and grandchildren, voluntarily consented to undergo a detailed medical history, physical
examination, and medical tests every three to five years, creating a wealth of data about physical
and mental health, especially about cardiovascular disease. The study found high blood pressure
and high blood cholesterol to be major risk factors for cardiovascular disease).

Page 2 of 7
8 November 2023

Advantages:
1. Design of choice for studying cause, course, and risk factors of a disease.
2. Since a probability sample is taken from the reference population, it is possible to
generalize from the sample to the reference population with a known degree of certainty.
3. One can test hypotheses of casual relationship
4. Since a cohort study starts with two groups of people, all of whom are free of the
condition and who are observed for a specified period of time; incidence of disease in
both groups can be calculated (Figure 2)
5. It is possible to quantify the risk of developing the condition, both relative risk (RR) and
attributable risk can be calculated from cohort studies.
6. Temporal sequence between exposure & disease can be more clearly established.
7. Particularly suited for assessing the effects of rare exposures.
8. Study multiple outcomes related to a specific exposure.

Figure 2. Calculation of the incidence in cohort study

Disadvantages:
1. They are very costly in time, money, personnel, efforts, and patient follow-up.
2. As long-term studies, they are rarely feasible for studying rare conditions.
3. Cannot be used to prove causation.
4. The most serious problem is the problem of attrition or loss of people during the course of
the study, because of migration, death, or refusal to continue to participate in the study.
5. There may also be attrition among investigators who may lose interest, leave for another
job, or another project.
Retrospective (Historical) cohort study
In retrospective cohort studies, two groups are retrospectively identified and “prospectively”
compared according to the following model: A cohort of healthy subjects is subdivided into
two groups – one exposed to a given factor and the other nonexposed to the same factor

Exposure * Start of the study outcome

While in prospective cohort study
* Start of the study Exposure outcome
In a prospective cohort study, the investigators are present from the beginning to the end of
the observation period. However, through the use of a historical or reconstructed cohort study, it
is possible, under special circumstances, to have many advantages of a cohort study without the
continuous presence of investigators.
The design of this type of study depends upon the availability of data from which to reconstruct
Page 3 of 7
8 November 2023

cohorts of persons with a characteristic, and of those without the characteristic, from medical
records and to follow their records (or to follow up the patients themselves) at a later time to see
which have developed a condition and which have not. In other words, although the investigator
is not present in the early period of the study, he reconstructs a reference population from records
and then proceeds as though he had been present throughout the study.
The advantages of retrospective cohort studies are that they are less expensive to perform than
cohort studies and they can be performed immediately because they are retrospective. Also due to
this latter aspect, their limitation is poor control over the exposure factor, covariates, and potential
confounders.
2 Case-control (case history or retrospective) study
For case-control studies (also known as retrospective studies or case-history studies), a group of
affected persons is compared with a suitably matched control group of non-affected persons.
For example, in a study designed to test the hypothesis that cigarette smoking is an important
factor in causing lung cancer; a number of patients with this disease (cases) were questioned
about their smoking habits. Similar questions were asked of a group of patients who had cancer at
other sites (controls). This enquiry showed significant differences in the smoking habits of cases
compared with controls. It was a case-history study in that the subjects were selected on the basis
of being affected or non-affected persons.
In case history study sample chosen on the basis of outcome (cases), plus comparison group
(controls). It starts with people who have a condition and then investigates them retrospectively to
see whether or not they have (or have had) the characteristic and compared with a group or groups
of controls who are free of the condition.
The basic format of the study is shown in Figure 3, along with the example of oral cancer and betel
chewing Figure 4

Figure 3. case control studies

.

Page 4 of 7
8 November 2023

Figure 4. Case control study

Advantages:
1. It is a short-term (quick) study, easy to conduct and is relatively least expensive, requiring
much less time and expense, fewer personnel, and a smaller sample than a cohort study
2. It is the only practical approach for rare diseases (but not rare exposures)
3. It can support (but not prove) casual hypotheses. It is good for investigation of a
preliminary hypothesis (and multiple risk factors can be explored)
4. It establishes association (odds ratio can be estimated from it)
5. There is no problem of attrition (except when follow-up requires periodic investigations
and some subjects refuse to continue to cooperate).
6. When case history studies are repeated, and the results of the studies, especially those by
different investigators and in different population groups confirm each other, considerable
confidence can be attached to the conclusions. For example, the relationship between
smoking and lung cancer was demonstrated with case history studies before cohort studies
were conducted.
7. Sometimes it is possible, in case history studies, to demonstrate a dose-response
relationship, which adds to confidence in the study conclusions. For example, it was
Page 5 of 7
8 November 2023

possible to indicate through case history studies that the number of cigarettes a person
smoked per day affected his risk of getting lung cancer.
3 Disadvantages:
1. In a case history study, the universe or reference population is not known, and it is
therefore impossible to generalize from the sample to the universe,
2. Cannot assess incidence or prevalence of disease
3. It is impossible to make precise risk statements.
4. Bias is more likely to affect case history studies than cohort studies through selection,
selective recall, false labeling, or reverse bias.
5. Difficulty in finding matching controls
a. No. should be 1-4 times.
b. Matching (particular & conceptual problem).
c. Randomization (problem of specific control).
d. Types of control group
6. Selective survival can present a serious problem in case history studies. For example, in
the case of a highly fatal disease, such as septicemia, it might be found that most of the
cases come from high social classes; leading to the erroneous conclusion that septicemia is
more frequent among the well-to-do than among the poor. In fact, the opposite may be
true, and septicemia may be more frequent among the poor; but because of malnutrition
and lack of medical care, poor people with septicemia may die more often than well-to-do
people with septicemia. When a retrospective study is initiated, only those who have
survived will be available for investigation, thus creating an artificially high representation
of well-to-do and well-cared-for individuals.
4 Case-control versus cohort design
1. Case-control study
 Subjects selected on the basis of disease

 Analysis compares frequency of exposure in the two groups

2. Cohort study
 Subjects selected on the basis of exposure

 Analysis compares incidence of disease in the two groups

Figure 5. Figure comparing the design of case-control study with that of cohort study

Compared with cohort studies, case-control studies have the advantage of being relatively quick,
easy and relatively inexpensive. A significant number of cases can be assembled for the case-
Page 6 of 7
8 November 2023

history study and a variety of hypotheses can be rapidly screened. The more promising theories
can be further examined by the more laborious, time-consuming and expensive cohort studies.
The latter have the advantage of giving a more direct estimation of the risk from exposure to each
factor.

Page 7 of 7