Experimental Design of Cosmetics Human Efficacy Evaluation

Song Yanqing, Ma Xue, Pan Yao, Zhao Hua

Department of Cosmetics, School of Science, Beijing Technology and Business University, China

Abstract

Key words

cosmetics; efficacy evaluation; clinical study; experimental design

With the improvement of the regulatory requirements of cosmetics market in China, the efficacy claim of cosmetics needs to provide sufficient scientific basis, and the evaluation of cosmetic efficacy is an effective means to support the claim of efficacy.Among the various experimental methods of efficacy evaluation, clinical trial, as the closest experimental method to the actual use of cosmetics, can most directly and truly reflect the efficacy of cosmetics, playing an irreplaceable role in cosmetic efficacy evaluation.The scientific and rigorous experimental scheme design is the key to improve the authenticity of the research results.It is helpful for human research to provide more accurate and reliable research data by the detailed elaboration of the basic principles, common design types and key points of the experimental design of human efficacy evaluation.

Experimental design of human efficacy evaluation of cosmetics

Cosmetic human efficacy evaluation is an experimental study to evaluate the function and effect of cosmetic products in a specific population.In order to make the whole evaluation go on smoothly and collect complete and reliable data, it is necessary to arrange the experiment scientifically and reasonably before carrying out the research, which is called experimental design.[1]The experimental design includes two aspects: professional design and statistical design.The professional design mainly includes the selection of research topics, the selection of research objects, the determination of observation indices, the adoption of relevant detection techniques and methods, and the consistency of operational standards and decision standards and so on.[2]

Basic elements of experimental design

Subject.The subjects are the objects of the experimental factors.In cosmetics human efficacy evaluation, it refers to the subjects who participate in the clinical trial.The subjects in the human efficacy evaluation experiment should be selected not only according to the purpose and professional knowledge of the study, but also taking into account the general conditions such as race, sex, age, health status, skin condition, as well as some social background, including occupation, hobbies, living habits, living conditions and psychological status.The differences of the above factors will have an impact on the experimental results.Before the experiment, the quality standards of the subjects must be established, which usually refers to the inclusion criteria and exclusion criteria.At the same time, in order to ensure the accuracy of statistical analysis of the data, the appropriate number of objects is estimated by the specific conditions of the experiment.

Experimental factors.Some of the experimental conditions to be imposed or observed according to the purpose of the study are experimental factors, also known as “treatment”.[3]In cosmetics human efficacy evaluation, the experimental factors refer to the cosmetic samples given to the subjects.In the experimental study, in addition to the experimental factors, the subjects' own conditions will also affect the results, which is called non-experimental factors.In the experimental design, in order to more accurately evaluate the real impact of experimental factors on the results, it is often necessary to find out the key non-experimental factors according to the professional knowledge and experimental conditions and to control and arrange them reasonably so as to eliminate their impacton statistical analysis.

Experimental effect.The experimental effect is the effect and reaction of the experimental factors on the subjects.It is reflected by the observation index and is the core content of the experimental research.In cosmetics human efficacyevaluation, the experimental effect is the improvement of the skin of the subjects after the use of cosmetics products.The commonly used observation indicators can be divided into objective indicators and subjective indicators.The objective indicator is the quantitative observation index obtained by the skin noninvasive measuring instrument.Subjective indicators include visual assessment and subject self-evaluation.Visual evaluation is usually carried out by experts or doctors to qualitatively or graded the skin condition indicators of the subjects.The subjects' self-assessment is mostly carried out in a form of questionnaires.The subjects score their skin condition according to their personal subjective feelings.Although some indicators are objective, the judgment is affected by subjective factors, such as different doctors look at the same subject's facial photos to evaluate their skin condition.Such indicators can be called semiobjective indicators.

Basic principles of experimental design

Randomization principle.The randomization principle runs through the whole process of the design and implementation of the experiment.It is mainly reflected in the following three aspects in the cosmetics human efficacy evaluation: 1) In order to ensure the representativeness of the subjects, it is necessary to make any individual in the population have the same opportunity to be selected as a subject in the sampling.2) Any individual in the subjects has the same opportunity to be assigned to any treatment group to ensure comparability among the groups, that the distribution of non-treatment factors in each group are consistent.3) Any individual in the subjects has the same chance of being treated sequentially to balance the effects of the experimental order.Because the recruitment of the subjects is limited by objective conditions and maneuverability, the human efficacy evaluation experiment is usually not random in the actual sampling process.It is difficult to randomly select the subjects from the whole population, but the grouping of the subjects is random.[4]

Contrast principle.In the study of human efficacy evaluation, we must make a comparison to find out whether the cosmetics are good or not.Only when there is comparison can there be identification, and the establishment of a control is the basis for comparison.Because there are many non-experimental factors affecting the experimental results in a study, setting up a control group can make the experimental factors and the other factors between the two groups in the same state and make the non-experimental effects comparable.Thus the difference of the outcome indices between the two groups can truly reflect the magnitude of the experimental factors.Some skin conditions will naturally alleviate, and only the establishment of a control can judge the relationship between the use of cosmetics and natural outcome.The function of the control is to improve the identification ability of the experiment and make the conclusion more convincing.

Replication principle.“Replication” usually has two meanings in cosmetics human efficacy evaluation, that is, repeated sampling and repeated measurement.Repeated sampling is to measure an objective index on the same subject at the same time and repeated detection several times to eliminate the error caused by the operation method.Repeated measurement is that the subjects observe many times in chronological order and measure the index every other time.The purpose of repeated measurement is to observe the dynamic change of the objective index with time.The main function of the replication principle is to control and estimate the random errors in the experiment.[5]

Blind principle.The authenticity of the research information of cosmetics human efficacy evaluation is easily affected by the subjective factors of the subjects and researchers, which come from the experimental design stage and the data collection or analysis stage.Blind method can be used to overcome this bias.Blind is a strategy to avoid knowing what to do with the subjects.That is to keep the grouping information of each subject confidential during cosmetic use and follow-up, so that the participants in the study, includes subjects, observers or outcome evaluators (dermatologists or relevant professionals), data collectors and statistical analysts, are not aware of the subjects' grouping.[6]

Common experimental design types in the study of cosmetics human efficacy evaluation

The type of experimental design refers to a combination of experimental factors and their levels in a specific study or a form of structure.Among them, “level” refers to the different values of the strength or quantity of each experimental factor.[7]Researchers should choose the appropriate design type combined with the professional requirements according to the research purpose and the actual conditions of the research.The following is a brief introduction to several types of experimental design commonly used in the study of cosmetics human efficacy evaluation.[8,9]

Single design

Single design is to randomly assign all the subjects of the same population to each level group of experimental factors, receive different treatments and then compare their effects.Because there is only one experimental factor examined in the research design, it is called single-factor design.The level number of the experimental factors investigated is generally determined according to the purpose of the study.When the level numberk=2, it can be divided into paired design and group design; whenk≥3, it is a single factor multi-level design.The details are as follows:

Paired design.In the study of cosmetics human efficacy evaluation, the commonly used paired design type is self-pairing.Values of the same index are detected from the same area of each subject before and after the use of cosmetics, and these two values form a pair from each individual.In the process of testing and analyzing the clinical efficacy of ginseng anti-aging products, subjects were asked to use ginseng anti-aging products daily for 8 weeks by Li Zi'an[10]et al.The patients were followed up once in week 0, week 2, week 4, week 6 and week 8, and the related evaluation indices were detected.The anti-wrinkle effect of the product was evaluated by comparing the difference of each index at different time points before and after the use of the product.In addition, the same subject uses two different cosmetics in different parts of the body to compare the effects of these two treatments, which is another kind of paired design.In a comparative study on the moisturizing effect of two facial masks, the subjects were asked to use mask A and mask B on the left side and right side of face respectively for 2 weeks by Ma Yanfeng[11]et al.The moisturizing effects of the two masks were compared and evaluated between the left and right side of face.

Group design.Group design, also known as completely random design, is a random selection of two groups of subjects from the same population.The subjects receive two different treatments and then their effects are compared.During the clinical observation of the efficacy and safety of resveratrol external application, 40 female subjects with facial photoaging were randomly divided into two groups by Jia Lili[12]et al: One group was treated with resveratrol and the other group was treated with matrix.The whitening effect of resveratrol was evaluated by comparing the differences of each measurement index between the two groups.

Single factor multi-level design.Single factor multi-level design is to randomly divide subjects from the same population into three or more groups, each using different cosmetics with similar efficacy.In the human efficacy evaluation of four moisturizing skin care products, Xie Wenjian[13]et al.divided the subjects into four groups, one product in each group to evaluate the moisturizing effect of these four different skin care products.

Repeated measurement design

The repeated measurements are quite common in cosmetics human efficacy evaluation, which is the same index measured many times at different time points after usingthe cosmetics.Repeated measurement design can be used to understand the trend of the observed data changing with time in the course of the experiment.During the preliminary study on the immediate effect of a moisturizing mask, Ma Yanfeng[14]et al.asked the subjects to remove the mask after applying for 15 min and to determine the moisture content, melanin index and elasticity of the skin at 0, 1, 3 and 6 h, respectively to understand the changing trend of the three indices with time.

Key points of experimental design for cosmetics human efficacyevaluation

Define the purpose of the study and select the appropriate subjects

The human efficacy evaluation experiment is mainly used to evaluate the efficacy of a cosmetic product.That is, it is used to answer the question of whether cosmetics are effective and whether their claims are true.The general human experimental research contains four main contents: the subject, the experimental factors, the comparison method and the expected outcome.The essence of the research is to accurately consider, define and explain these four aspects in detail.[15]For example, if you want to study the spot removingefficacy of a cream, we can determine the purpose of the study as: to study the effect of spot removing (expected outcome) in the population with facial spots (subjects) before and after the use of a cream (experimental factors).Generally speaking, it is incomplete to regard “study the spot removingefficacy of a facial cream” as the purpose of the study.A precise purpose of the study can provide a prerequisite for deciding the specific methods to solve the problem in the next step.

The research subjectsare determined by the purpose of the study.The selection of the research subjects is defined and limited by the inclusion criteria and exclusion criteria.Therefore, reasonable inclusion and exclusion process is the basis to ensure the scientific nature and smooth development of human efficacy evaluation experiments.[16]The section “selection of subjects” is included in the relevant chapters of the evaluation and testing methods of human safety and efficacy in theSafety and Technical Standards for Cosmetics(2015 Edition),[17]and different inclusion and exclusion criteria are listed for different research purposes.It provides a good reference and guidance for researchers.

Establish a reasonable comparison

In the study of cosmetics human efficacy evaluation, we must adopt strict and reasonable control to eliminate bias to correctly evaluate the efficacy of cosmetics.So that researchers can make a correct evaluation.The commonly used control methods include blank control, standard control, selfcontrol and experimental control.[18]

Blank control refers to the control group without applying any treatment factors, that is, the control group does not use the tested cosmetics and maintain a natural state.The standard control, also known as the positive control, is based on the cosmetics claimed the same efficacy accepted by the industry as the control or using the competitive products for control.Self-control is that the control and treatment are carried out on the same subject.It can be divided into two situations: 1) The control area and the treatment area are set up in different parts of the subject's body.2) The comparison between before and after the use of cosmetics is also a self-control.Experimental control, also known as matrix control, refers to the control group applying the same basic formula as the treatment group without the functional component.Because the results are easily affected by experimental factors and other non-experimental factors, the establishment of experimental control can be used to eliminate the interference of a specific baseline treatment.[19]

The evaluation indices should be objective, reliable and easy to measure

Evaluation index is a ruler to measure the effect of experimental factors.It must be considered comprehensively in many aspects, such as the nature and characteristics of experimental factors, instruments and technical level.The selection of evaluation indexdetermines the success of the study.[20]It must be intrinsically related to the purpose of the study and can accurately reflect the effects of experimental factors.In addition, the evaluation index should also have a certain degree of objectivity, specificity, sensitivity and accuracy.Objective indicators should be used as far as possible, semi-objective or subjective indicators can be used as a supplement.For some semi-objective or subjective indicators, it is necessary to specify strict criteria for reading the results in advance and deceive the graders so that they do not know the grouping of the subjects.It is necessary to conduct unified technical training to minimize the influence of subjective factors on the judgment of the results.

Scientific estimation of sample size

The cosmetics human efficacy evaluation experiment belongs to the sampling study, which selects the appropriate representative individuals from the target population to do the study instead of the whole one.Based on the hypothesis testing principle, small sample size means low study accuracy, which has little practical value, while large sample size will cause unnecessary waste of resources and may also increase the opportunity of bias.Therefore, it is necessary to determine the appropriate sample size according to different design requirements.[21]In statistics, a sample size of more than 30 in each group can be regarded as a large sample.[1]The number of qualified subjects required by the Safety and Technical Standards for Cosmetics (2015 Edition)for human safety testing methods is at least 30.This is consistent with the minimum number of cases in clinical pharmacology and human safety evaluation trials of drugs regulated by theProvisions for Drug Registration.It can be seen that the sample size estimated by cosmetics human efficacy evaluation experiment is reasonable for at least 30 people in each group.Truly, the estimated sample size is the minimum requirement for theresearch.In practical work, considering the subjects'poor compliance, mid-way drop-off and loss of follow-up, it is generally possible to increase the sample size by 10% or less by 20% on the basis of the estimated sample size.[22]

Strictly abide by ethical and moral requirements

Ethics is a kind of moral norm used to distinguish the right and wrong moral concepts of human activities.Researchers must strictly abide by ethics in the whole process of human experimental research.Ethical design is an indispensable part of all human experiments and occupies an extremely important position.Ethical design should follow the principles of the Declaration of Helsinki, namely, impartiality, respect for personality, maximizing the benefits of the subjects and avoiding harm as much as possible.[23]Specific measures include informed consent, confidentiality of data, measures to reduce injury and to protect the rights and interests of subjects.

Group the subjects randomly

The comparability of the factors affecting the outcome between groups is the premise of accurately estimating and comparing the efficacy of the tested cosmetics.To achieve comparability between groups, the grouping process must have nothing to do with any known and unknown factors that may affect the results.This grouping method is often known as randomized grouping.There are two kinds of randomized grouping methods commonly used in cosmetics human efficacy evaluation experiment as follow.

1) Simple randomized grouping: The subjects can be divided into groups by coin tossing (the head and tail sides are designated as the experimental group and the control group, respectively), drawing lots or throwing dice etc.It is more scientific and reliable to use random numbers to group the subjects.Random numbers can be obtained from relevant statistical books or generated using statistical software.[24]

2) Stratified randomized grouping: According to the characteristics of the subjects that are the important non-experimental factors (such as age, sex, skin condition, etc.) affecting the experimental process and results, the subjects are first stratified and then randomly assigned to the treatment group and the control group in each layer.Usually, the selection of stratified factors needs to be determined on the purpose of the experiment and combined with professional knowledge.

In practice, researchers often misunderstand and misuse the randomized grouping method.The most common situation is to alternately assign subjects to different study groups refer to their date of birth (oddeven year), the end number of ID (odd-even number), the enrolled time (odd-even day) or the enrolled order (odd-even number).Because in these cases, the researcher can predict which group the subject will be divided into before the grouping.If the researcher is inclined to divide the subjects into a particular group, he may take steps to intervene in the grouping of the subjects.Once the researchers intervened the grouping, the randomization principle was broken.Therefore, alternating grouping is not a randomized grouping method in the strict sense, but belongs to pseudo-randomized grouping or quasi-randomized grouping.At the same time, randomized grouping is not ad libitum grouping.According to the list of subjects, the first half is divided into the treatment group and the second half is the control group, or the members of the treatment group and the control group are selected as the researchers' own wish, all of which are ad libitum grouping.This grouping method does not give subjects the same opportunity to enter different study groups, so this ad libitumgrouping should be avoided.

Application of blind method

To a certain extent, blind method is helpful to reduce the bias caused by subjective factors in the implementation of human efficacy evaluation research, so as to maintain the comparability among groups.According to the different degree of blindness, the blind method can be divided into single-blind method, doubleblind method and three-blind method.[25]Single-blind method and double-blind method are widely used in the study of cosmetics human efficacy evaluation.The single blind method means that the subjects know nothing about which group they are in, while the researchers know, that is, only the subjects are blind.Single-blind method can avoid bias caused by subjective factors from the subjects.In order to minimize the subjective, psychological factors, the surrounding environment and other factors impacting the evaluation results, neither the subjects nor the researchers who give cosmetics or analysis the results know the grouping of the experiment in the doubleblind method, and the study designer arranges and controls the entire experiment.The experiment without blindness is called open experiment, that is, the subjects and researchers all know the grouping.The non-blind design is suitable for experiments with objective observation indices.

Improve the compliance of subjects and reduce the loss of follow-up

Compliance refers to the extent to which subjects use cosmetics correctly and in sufficient quantities as required by the researchers.[26]Compliance is the premise to achieve cosmetic efficacy.If subjects do not use the product, the study cannot evaluate the true efficacy of the product.Reduced compliance usually results in underestimating the efficacy of cosmetics, so compliance is very important in human efficacy evaluation.There are many forms of non-compliance, including insufficient use of products, misuse, complete non-use, plus other products with the same efficacy and so on.The loss of follow-up was the loss of contact between the researcher and the subject.The withdrawal is still connected but the subject is unwilling to continue the study.Their impact on the study is the same, resulting in the loss of subjects and the loss of outcome data.Therefore, loss of follow-up includes withdrawal.Loss of follow-up is inevitable in human experimental research.Researchers should fully understand the consequences of it to the study and take measures to minimize its occurrence.[27]

Record adverse reactions and evaluate the safety of test samples

Adverse reactions are adverse medical changes after the use of cosmetic products.The occurrence of adverse reactions is a variable used to evaluate the safety of cosmetics and it is also the data that must be collected in each cosmetics human efficacy evaluation experiment.During follow-up, it is necessary to ask and observe the occurrence of adverse reactions and judge the relationship of adverse reactions occurrence and the use of cosmetics according to professional knowledge, and then make a detailed record.[28]The common adverse reactions associated with cosmetic use are mainly irritating and allergic reactions to local skin, including erythema, infiltration, papules, edema, blisters, bullae, pruritus, tingling and burning.[29]What needs to be recorded includes a description of the symptoms, the time and duration of occurrence, the severity, the examination, treatment and the final results of adverse reactions and how the researchers handle the situation.

Implement the standard operating procedures to ensure the standardization of the research practice

Cosmetics human efficacy evaluation is the most persuasive method to provide scientific and objective basis for cosmetic efficacy claim.Quality control is an important means to ensure the results are scientific, reliable and true throughout the experimental process.In order to guarantee the standardization of drug clinical trial process, International Conference on Harmonization of Technical Requirements for Registration of Pharmaceuticals for Human Use (ICH) issuedGuideline for Good Clinical Practice[30]and China also promulgate theGood Clinical Practice(GCP).[31]These guidelines and specifications provide a good reference for the quality control of cosmetics human efficacy evaluation experiments.GCP requires all parties involved in clinical studies to develop standard operation practice (SOP) and comply with them.SOP refers to the development of uniform, detailed, standard operating procedures for the completion of a specific job responsibility in a clinical trial in order to achieve uniformity.[32]The results obtained by different operators or the same operators at different times following the SOP can be almost the same.

Conclusion

The research on cosmetics human efficacy evaluation can reflect the actual effect of cosmetics more accurately than animal experiments, cell-based experiments and consumer surveys, and provides the most valuable information for the efficacy claim of cosmetics.A perfect experimental design of cosmetics human efficacy evaluation needs to arrange various experimental factors reasonably, correctly estimate the sample size, strictly control the experimental bias and effectively analyze the experimental data so as to use less resources and time to maximize the scientific and reliable basis for the efficacy claim.Therefore, when designing a human efficacy evaluation research, it is necessary to ensure that the experimental design meets the requirements of science, preciseness, order, economy, efficiency and perfection.Strict quality control should be carried out in the implementation stage to reduce bias and improve efficiency.