Course catalogue doctoral education - HT17

  • Ansökan kan ske mellan 2017-04-13 och 2017-05-15
Application closed
Title Design and analysis of twin and family-based studies
Course number 2893
Program Epidemiologi
Language English
Credits 1.5
Date 2017-10-23 -- 2017-10-27
Responsible KI department Institutionen för medicinsk epidemiologi och biostatistik
Specific entry requirements Epidemiology I: Introduction to Epidemiology, Biostatistics I: Introduction for epidemiologists, Epidemiology II: Design of epidemiological studies, Biostatistics II: Logistic regression for epidemiologists and Biostatistics III: Survival analysis for epidemiologists or corresponding courses
Purpose of the course This course focuses on potential designs and analyses using twin- and family-data. Methods to estimate within-family associations and heritability are covered.
Learning outcomes After successfully completing this course you as a participant are expected to be able to:

- discuss the difference between a within-family analysis and a more standard (e.g. between-family) statistical analysis,
- select an appropriate within-family/heritability analysis for a given dataset, based on a specific research question,
- discuss how to perform within-family/heritability analyses using the statistical software R,
- interpret the output from a within-family/heritability analysis, and compare with a more standard statistical analysis,
- discuss assumptions made in heritability analysis, and how violations may affect the results.
Contents of the course The aim of empirical research is often to estimate the causal effect of a particular exposure on a particular outcome. A complicating feature of observational studies is that the exposure-outcome association is typically confounded, and cannot be given a causal interpretation. The standard approach to deal with confounding is to control for confounders in the analysis, e.g. by regression modeling. However, many confounders may be difficult to measure, or unknown to the investigator. An appealing solution is to study within-family associations, which are automatically controlled for all factors that are shared within the family (e.g. socioeconomic status, genetic factors). In this course we will focus on the theory and practice of within-family analyses. In many studies, the research question is to what extent a phenotype is caused by genetic factors. Frequently though, there may be no obvious candidate gene, and financial limitations may prohibit a genome wide scan. An appealing solution is to study whether the phenotype tends to run in families; the stronger genetic influence, the larger familial heredity. A commonly used design to estimate the fraction of variation in an outcome which may be attributable to genes and environment is the classic twin methodology. In this course we will cover the concept of heritability, its underlying assumptions, and applications in the classic twin method. Within-family analysis and bivariate heritability analysis (i.e., quantitative genetic analysis of two phenotypes) complement each other. Although within-family analyses require fewer assumptions, bivariate heritability analyses may yield additional information. In this course we will compare and contrast the methods.
Teaching and learning activities Different strategies for teaching and learning, such as interactive lectures, small group discussions and exercises on selected topics, will be used.
Compulsory elements The individual, oral examination.
Examination An individual, oral examination will take place the last day of the course. Each student will present a hypothetical study in which all the intended learning outcomes should be addressed. The examination will be performed in small groups with one examining teacher in each group. Feedback from peers will also be emphasized. Students who do not obtain a passing grade in the first examination will be offered a second examination within two months of the final day of the course. Students who do not obtain a passing grade at the first two examinations will be given top priority for admission the next time the course is offered. If the course is not offered during the following two academic semesters, then a third examination will be scheduled within 12 months of the final day of the course.
Literature and other teaching material Suggested reading: Allison PD. (2009). Fixed effects regression models, Quantitative Applications in the Social Sciences, Vol. 160. SAGE: Los Angeles. Neale & Maes (2004). Methodology for Genetic Studies of Twins and Families. Kluwer Academic Publishers B.V. Dordrecht, The Netherlands. Available at
Number of students 8 - 25
Selection of students Eligible doctoral students, with required prerequisite knowledge, will be selected based on 1) the relevance of the syllabus for the applicant's doctoral project (according to written motivation), and 2) date for registration as doctoral student (priority given to earlier registration date). To be considered, submit a completed application form. Give all information requested, including a description of current research and motivation for attending, and an account of previous courses taken.
More information
Course directors Teachers will be: Ralf Kuja-Halkola, PhD, Department of Medical Epidemiology and Biostatistics, Karolinska Institutet. Arvid Sjölander, PhD, Associate Professor, Department of Medical Epidemiology and Biostatistics, Karolinska Institutet. Paul Lichtenstein, PhD, Professor, Department of Medical Epidemiology and Biostatistics, Karolinska Institutet. Brian D´Onofrio, PhD, Professor, Department of Psychological and Brain Sciences, Indiana University, Indiana, USA. The course is given in collaboration with the Swedish INterdisciplinary Graduate School in register-based research (SINGS), link:
Earlier evaluation of the course Evaluation report
Course responsible Ralf Kuja-Halkola
Institutionen för medicinsk epidemiologi och biostatistik
Contact person Gunilla Nilsson Roos
Institutionen för medicinsk epidemiologi och biostatistik
08-524 822 93