Study Designs

Question 1

what are the main research study designs

Answer 1

1. Laboratory analyses
2. Animal models
3. Studies with consumers
4. Epidemiology
5. Randomised Controlled Trails

Question 2

what are the advantages of laboratory analyses

Answer 2

1. Most laboratory analyses can be easily repeated (replicated).
2. Many validated standard methods are available that use suitable automated equipment, calibration standards, control samples and Standard Operating Procedures.
3. In vitro methods, unlike biochemical or other analyses can give physiological data, and are usually much more rapid and easier to use and to control than animal models or human studies. In vitro methods can be used to screen (evaluate) large numbers of compounds / substances.

Question 3

what are the limitations of laboratory analyses

Answer 3

1. All samples need to be correctly stored before analysis.
2. With food samples, variations due to other factors, such as origins and processing methods need to be considered, if these are not the subject of the study. -
3. Are the methods measuring the biologically relevant components such as vitamins, or co-factors?
4. Do in vitro methods really mimic the in vivo situation? Are the concentrations used in the same range as those found in vivo?
5. Validated methods are not available for all potential analytes.

Question 4

what are the pros of animal studies

Answer 4

1. These are much easier to control than human studies.
2. All tissues can be accessed, and lifetime studies can be carried out with animal models.
3. Animals are more complex models than in vitro, but are still not humans, therefore interpret with caution.

Question 5

what are the studie types with consumers and populations

Answer 5

-Consumption and consumption patterns (dietary / food / nutrients etc) at individual, group or population levels. These are often similar to descriptive epidemiological studies, also often used in marketing surveys and in health promotion studies
-Behavioural studies, for example understanding motives behind food purchasing and consumption patterns. These are often linked to marketing surveys and to health promotion initiatives.
-Sensory evaluations used in product development and in research. There is a wide range of tests available for assessing sensory qualities (sensory profiling; difference tests etc)

Question 6

what are the pros of consumer/population studies

Answer 6

1. Use humans.
2. Can use “real life” settings.
3. Can use validated “instruments” (questionnaires).

Question 7

what are the limitations of consumer/population studies

Answer 7

1. May be large variation due to many factors, e.g. biological (gender, age etc), psychological (attitudes, beliefs), socioeconomic (income, ethnicity).
2. If these factors are not the subject of the research, then need to analyse data using suitable statistical methods.
3. Often need very large sample sizes

Question 8

what is the gold standard test

Answer 8

RCT

Question 8

what are descriptive studies

Answer 8

these describe the incidence of disease and its relationship to other factors. Such studies generate hypotheses.

Question 9

what are analytical studies

Answer 9

these aim to test hypotheses about the causes of diseases.

Question 10

what are the types of observational studies

Answer 10

1. Time trends
2. Cross-sectional studies – give data for a population group at one point in time.
3. Cross-cultural studies – compare data from different groups (e.g. countries, regions, ethnicity, migrants, lifestyle, religion).

case control + cohort

These studies can be applied to individuals, but generally use data for the whole group.

Question 11

what are the pros of observational studies

Answer 11

Relatively quick and inexpensive (can use data already collected). Can be used to give early indications of relationships

Question 12

what is a limitation of observational studies

Answer 12

Groups can be very heterogenous, thus the use of group (population) data may not apply to individuals.

Question 13

what is a case control study

Answer 13

the disease (cases) are compared with individuals without the disease (controls)- then exposed and unexposed

It is essential that the controls are matched with the cases for demographic variables.

Both disease data and dietary data are needed. The dietary data is usually collected retrospectively (looking back).

Question 14

what are the pros of case control studies

Answer 14

1. Relatively inexpensive
2. Can test hypotheses quickly
3. Since the focus is cases, can be used with rare diseases
4. Can test a number of hypotheses / risk factors at the same time.
5. Matching cases and controls can help eliminate confounding factors

Question 15

what are limitations of case control studies

Answer 15

1. Need to take care in selecting cases (ensure diagnosis is correct; account for possible dietary and lifestyle changes, and early mortality in cases).
2. Need to ensure that controls are well-matched for demographics, but free of the disease.
3. Needs information from cases and from controls on previous diet and lifestyle
4. Retrospective in nature, recall bias.

Question 16

what are cohort studies

Answer 16

A cohort is a defined group of subjects they are prospective (look forward).

A defined population group, free of overt disease, is assessed for dietary and lifestyle factors, and followed up over a period (usually many years) to evaluate the development of disease.( exposed and unexposed and then controls and cases)

The dietary and lifestyle factors of those with the disease are compared with those who do not develop the disease.

Question 17

what are pros of cohort studies

Answer 17

1. Avoid many of the limitations associated with case-control studies.
2. Using populations gives a good assessment of diet-disease relationships.

Question 18

what are limitations of cohort studies

Answer 18

1. Often need very numbers and a long time.
2. Need to stay in contact with subjects over the period of the study.
3. Lifestyle and diet may change over time (need to re-assess)
4. Very expensive.
5. Not suitable for rare diseases.

Question 19

what are the types of intervention studies and what do they describe

Answer 19

Experimental trials; randomised controlled trials [RCT])
The types of study described above can indicate diet-disease links.

Question 20

what are RCT

Answer 20

in RCT’s the researcher intervenes to modify the dietary factor that is hypothesised to be linked to the disease / disease risk factor.

RCT’s which show that a dietary change can result in a change in the incidence or risk of a disease gives the strongest experimental support for a link.

Dietary intervention studies are similar in design to studies to test new drugs or treatments.

Question 21

what can intervention studies vary in

Answer 21

duration and end points

Question 22

what are the features of a RCT

Answer 22

Subjects are randomly assigned to receive either the test (treatment) or control (placebo).

The randomisation aims to ensure that the test and control groups will be very similar in all respects (age, gender, BMI, lifestyle, physiologically).
Thus any effects observed on the test, can be attributed to the treatment.

It is also important, to try to ensure that neither subjects nor researchers know who is on the test or the control. This is achieved by double-blinding

Question 23

what are pros of intervention studies

Answer 23

1. Best experimental method for studying diet-disease relationships.
2. Can give direct evidence of benefits, and provide strong evidence to support dietary change in the population.
3. It is possible to study more than one factor at a time and / or to do dose-response studies

Question 24

limitations of intervention studies

Answer 24

1. May be difficult to blind subjects or researchers in some food-based studies. 2. Difficult to control single factors in food-based interventions; thus, effects may not be due to the factor under investigation. 3. May take many years to get results. 4. Some diseases are multifactorial, and it can be difficult to assess more than one factor at a time. 5. The risk factors must be validated for the disease under study. 7. Ethical considerations.

Question 25

what are the 2 designs of intervention studies

Answer 25

parallel and intervention studies

Question 26

what are parallel studies

Answer 26

test and control groups are randomly assigned, but are independent. The groups are usually on the study at the same time in parallel a-> a b ->b

Question 27

what are cross over intervention studies

Answer 27

test and control groups are randomly assigned and independent are usually on the study at the same time in parallel after the first leg there is a washout period with no intervention after which the groups cross over treatments so that both groups receive test and control

Question 28

pros of parallel studies

Answer 28

1. Can be used for very long-term studies, and use “hard” endpoints (overt disease or death)   2. Take less time than cross-overs that test the same hypothesis.   3. There is no risk of carry-over. [Carryover is when the effects of the first period can affect the second period].   4. With food-based studies, blinding is easier, because subjects only receive one treatment.

Question 29

pros of cross over studies

Answer 29

 1. Using both test and control on both groups greatly reduces “between-subject” variation.   2. Increases statistical power, and reduces the number of subjects needed.

Question 30

what is a placebo

Answer 30

(Latin = I shall please) A placebo is a substance or a treatment that the subject or patient thinks is therapeutic. Placebos are not active, and are used as the control in trials.

Question 31

what is the placebo effect

Answer 31

A placebo effect occurs when the subject or patient responds positively to the placebo. Beneficial effects that are found with the placebo, are thought to be due to the power of suggestion (the subject thinks that the treatment is beneficial).

Question 32

what is easy and hard to blind

Answer 32

Easy to blind capsule/powders (vits etc), hard for whole foods/diets, appearance, taste and macronutrient balance

Question 33

why is blinding in studies important and who is blinded

Answer 33

This is to prevent possible bias (intentional or unintentional). For example, researchers may take more care when assessing test subjects, or analysing their samples. Researchers may also give unintentional signals to subjects so that they know if they are on test or control.   Thus, studies need to be blinded for both the subject and the researchers (double-blind).

Question 34

what does it mean that subjects are blinded

Answer 34

This means that subjects do not know if they are on the test treatment or the control (placebo). This is to take account of possible placebo effects.

Question 35

what does it mean for researchers to be blinded

Answer 35

This means that the researchers will not know which subjects are on the test treatment or the control (placebo).

Question 36

what studies are ideal

Answer 36

double blind trials

Question 37

what are double blind trials

Answer 37

Treatment and placebo are given codes which that are assigned to the subjects by people independent of the study. Thus, the researchers do not know who is on control or treatment.   The codes are stored independently. The code must only be broken when the trial has been completed, and all analyses finished.   Thus, the researchers cannot give different managements or subconsciously indicate if subjects are on control or treatment. Double-blind trials are the best, and should eliminate any placebo or researcher effects.

Question 38

what is triple blinding

Answer 38

Usually the double-blinding is includes anyone involved in subject management or diagnosis, such as radiography or pathology, and to statisticians evaluating the data. Sometimes this is called a “Triple-blind trial”. However, this is really only double-blinding of subjects, and all in the research team.

Question 39

what are single blind trials

Answer 39

In these, the researcher knows if it is placebo or treatment but the subject or patient does not know.   Since the subject does not know if they are on treatment or placebo, there should be no placebo effect.   However, because researchers know which subjects are on treatment or placebo, they may intentionally or unintentionally influence the results of the study.

Question 40

what blind studies can be difficult to do

Answer 40

Some trials, such as those that use distinct foodstuffs, can be difficult to double-blind.

Question 41

what are open trials

Answer 41

In open trials, both the researcher and the subjects or patients know the details of the treatment.

Question 42

why can open trials be biased

Answer 42

These trials may be biased, since they do not consider the placebo effect, and the researchers can influence the results. However, open trials are unavoidable sometimes, because it is not possible to design a placebo. For example, where a convincing placebo may be too dangerous or unethical (e.g. sham surgery), or where the subject needs to actively participate such as dieting or physiotherapy. Thus, it may not be possible to blind all studies.

Question 43

what are the theoretical aims of designing studies

Answer 43

1. Maximise any possible differences due to treatment (Maximise systematic treatment variability) 2. Control or reduce non-treatment differences (Control extraneous systematic variability) 3. Minimise differences due to errors in measurements or the way measurements are made, both biological and analytical) (Minimise error variability)

Question 44

what is the practical aim of designing studies

Answer 44

1. Provide valid answers to research questions (hypotheses) in a cost-effective and timely manner.

Question 45

what does * mean in a graph

Answer 45

significance

Question 46

what do multigenerational studies allow

Answer 46

access to all body organs

Question 47

what does qualitative tests allow

Answer 47

how you think and feel asking questions

Question 48

what is an example of case report

Answer 48

a doctor and patients interaction- non clinical rigor

Question 49

is a cohort retro or pro

Answer 49

prospective

Question 50

what does case control start with

Answer 50

start with sick and non sick- asks about past- which is more bias

Question 51

what does the placebo effect look at

Answer 51

physchological effect

Question 52

in a meta analysis question what should you mention

Answer 52

PICO and meta analysis