To the top of the page
Chapter 2.7.4
Module:  2.
Nutritional supplements bioactivity, functional properties and safety: in vitro & in vivo studies
Unit:  2.7.
Design of human clinical studies
Chapter:  2.7.4.
Analyzing the results of a human clinical study on a nutritional supplement

Analyzing the diet

As mentioned, the volunteers need to keep dietary records during the study, either acute or prolonged. Analysis of their diet will help you to validate that they followed the dietary plans or instructions given and that there are no unacceptable differences between measurements. Dietary records can be analyzed for energy, macronutrient, and micronutrient intake through the use of commercially available or custom-made software, based on existing food databases (more 2.7.16).

Excluding non-conforming volunteers

If some volunteers failed to follow important parts of the instructions, they must be excluded from the study (more 2.7.17). Data from such volunteers may not be valid and may distort the analysis.

Analyzing biological samples

If the study design included collection and storage of biological samples, it is preferable to analyze all of them together in the end to eliminate or reduce day-to-day variability and save time and money. Schedule the analysis by deciding how many samples you will analyze each day and what you are going to measure. Analysis must be done according to established laboratory methods and good laboratory practices to ensure validity (more 2.7.18).

Unblinding the study and assigning the data to treatments

When all measurements are completed, you must unblind the study (if it is a blinded one). Therefore, ask the person who did the randomization to give you the copies he/she has made. Then, assign each volunteer's data to the experimental or control treatment.

Carrying out the statistical analysis

When all data have been assigned to treatments, you need to perform the statistical analysis. Your basic question is this: "Do the effects of the experimental supplement differ from those of the control supplement?" However, there may also be interesting questions regarding, for example, gender differences or correlations between variables. Numerous statistical tests exist to answer these questions (depending on the design of the study and structure of the data), which cannot be covered here. If you are not familiar with statistics, we highly recommend seeking the assistance of an expert.

Reviewing the literature

Search authoritative bibliographic databases, such as those mentioned above, to find new relevant scientific papers that may have been published since your original search. Then, you will need to compare your findings with those of similar clinical studies and try to find what is novel, what is similar, and what is different (and why). At the end, draw the conclusions deriving from your findings.

Disseminating your findings

Congratulations, your study is now finished! However, you will need to disseminate your findings to let other people know. This can be done in several ways: