Aims/Hypothesis:

Twin data are valuable for quantifying genetic and environmental contributions to phenotypes, including both disease risk (e.g. type 2 diabetes) as well as quantitative traits (e.g. body mass index). However, twins can differ from singletons insofar as they tend to be born earlier (37 weeks rather than 40 weeks), and smaller (lower birth weight). The extent to which twins differ from singletons on a range of other early life characteristics (e.g. adiposity, growth, appetite and illness) is unclear. It is important to establish the extent to which twin data are comparable to singleton data, in order to ensure that results derived from twin studies may be generalisable to singletons.

The primary aim of this project is to establish the extent to which twins are similar to singletons on a variety of early life characteristics, by comparing twin data from two UK-based paediatric cohorts (the Twins Early Development Study, and Gemini - Health and Development in Twins), with singleton data from a number of large UK-based cohorts including ALSPAC. We hypothesize that twins will not differ from singletons on early life characteristics, thereby supporting the generalisability of findings from twin studies.

Our secondary aim is to test if the relationship between different early life characteristics is the same for twins and singletons.

Exposure variable:

We do not have exposure variables as such, as we are comparing summary statistics across different cohorts. 'Twinness' or 'singleton' (captured by the cohorts themselves) will be the primary 'exposures' of interest.

Outcome variables:

Our primary interest is anthropometric data. Primarily, we will compare anthropometric data from the singletons in ALSPAC to twin data from the Gemini cohort (at a number of different age points from birth to 5 years) and the Twins Early Development Study (TEDS) (at 7, 10, 11 and 12 years).

In addition to anthropometric variables, we will be comparing the twins and the singletons on appetitive data. In the Gemini cohort we have data from the Baby Eating Behaviour Questionnaire (BEBQ) and the Child Eating Behaviour Questionnaire (CEBQ) and would be looking for a comparator measure for appetite within ALSPAC. We also hope to compare feeding behaviour; encompassing breast versus bottle feeding to ascertain, for example, whether having multiple babies would make breast feeding more difficult. Other variables that we will examine are illness (to establish if the shorter gestational period of twins effects their general health in the early years of their life), infant temperament, physical activity and sleep.

One of the primary interests of the Gemini Study is to understand the predictors of weight gain in early life. As such, we have established relationships between appetite and weight, and sleep and weight, using the twin data. We will therefore examine if the magnitude of the association between appetite and weight, and between sleep and weight is the same for twins and singletons.

We are also planning to compare our twin data to singleton data from the Cambridge Baby Growth Study cohort and the Millennium Cohort Study.

Confounding variables:

When making comparisons between the cohorts we will take account of gestational age, postnatal age, gender, parental age at birth, parental health, ethnicity, and sociodemographic information.