Adoption studies are used to estimate the degree to which variation in a trait is due to environmental and genetic influences. There are three typical strategies that are used in adoption studies:

• Parent as proband: When the parent has a disorder or behaviour that is studied. In other words, if you are studying schizophrenia, you are looking at children who were adopted whose mothers have been diagnosed with schizophrenia.
• Adoptee as proband: When an adoptee develops a disorder or behaviour.  This is then compared to both the birth parents and the adoptive family. In other words, if you are studying schizophrenia, you are looking to see if either the adoptive or birth family has schizophrenia as a way of determining the origin of the adoptee's illness.
• Cross-fostering design: If the mother already has an illness before giving a child up for adoption, this may be a confounding variable.  In this design, one group of adoptees has ill parents and after adoption has been raised by healthy parents; the second group has healthy biological parents but has been brought up in a family where a parent has become ill.

Proband simply means - the person in the study who has the trait which is being studied.

It is important to note that although adoption studies may imply a genetic origin of a behaviour, research done before the completion of the Human Genome Project (2003) does not actually know the genotype of the adoptee and parents.  Today adoption studies are not as common as they used to be.  Today pschologists can do linkage studies and GWAS, allowing for a larger range of data and a more direct approach to studying the influence of specific genes on behaviour.

Sorensen et al (1992) studied 840 adult Danish adoptees. The study used the adoptees as probands. The height and current weight were measured to calculate BMI.  In addition, school records were obtained to determine their height and weight at ages 7 - 13 years old. Data from both the birth parents and the adoptive family were obtained through questionnaires, asking for height and current weight, as well as the maximum weight. This data was used to calculate BMI.

In addition, adoptees were given a questionnaire in order to document environmental factors which may have played a role in obesity.  This included the age of the child and parent at time of adoption, family income and smoking habits.

The results showed the following correlations across the ages 7 - 13 years.

Biological mothers: 0.17                Adoptive mothers: 0.10

Biological fathers: 0.16                  Adoptive fathers: 0.03

Biological siblings: 0.59                 Adoptive siblings: 0.14

When looking at the adult data, they found the similar levels of correlation, both for current weight and for maximum weight. None of the environmental factors studied showed any significant correlation with BMI. This implies a potential genetic influence.

The study has a large sample size, making the data more reliable.

The lack of a significant correlation between environment variables and BMI indicates the likelihood of a genetic root of the behaviour.

The study is based on school data that cannot be verfiied for accuracy.  However, the use of this data allows for a more longitudinal approach to the study.  The correlation was seen to be consistent over time.

The study makes use of BMI.  Many health psychologists argue that this is a not a good measure of overweight and obesity.

Adoption studies are often problematic because of selective placement - that is, often children are adopted into families that are similar to their birth family. 

As the genotypes of the children and their parents are not identified, it is not possible to know which genes - if any - are having an influence on their weight.