Our research
Back to listing

Environmental Risk (E-Risk) Longitudinal Twin Study

About the cohort

PI: Louise Arseneault as PI (Terrie E Moffitt and Avshalom Caspi study were study founders) 

  • Start date: 1999 (first data collection)
  • Age at recruitment: 5
  • Sample size at recruitment: 2232
  • Estimated current sample size: 2066

The Environmental Risk (E-Risk) Longitudinal Twin Study of childhood behaviour disorders was proposed in 1997 and funded for the period from 1998 to 2003. The E-Risk Study began by constructing a nationally representative 2-year birth cohort of 1116 families having 2232 5-yearold twin children born in 1994-95. The Study was planned as a 15-year longitudinal study concentrating on three developmental stages: (1) the transition to formal schooling, with assessments at ages 5 and 7 years, (2) the transition to secondary school, with assessments at ages 10 and 12, and (3) the transition out of school, with assessments at age 18. Comprehensive home visits were conducted when the children were age 5 (1999-2000) and age 7 years (2001-2002), retaining 98% of the cohort; again when they were age 10 years (2004-2005) and age 12 years (2006-2007) retaining 96% of the cohort; and most recently, at 18 years (2012–2014) retaining 93% of the cohort.

At age 18, E-Risk participants who did not participate in the study did not differ from those who did on initial age-5 measures of family socioeconomic status (SES), IQ scores, or internalizing or externalizing behaviour problems. Home visits at ages 5, 7, 10 and 12 years included assessments with participants and primary caretakers; the visit at age 18 included interviews only with participants. Each twin was assessed by a different interviewer.

The E-Risk Study was proposed to improve basic understanding of the etiology of mental disorder, and to inform prevention. Programmes to prevent behaviour disorders would be improved by: (a) compelling demonstrations that environmental risk factors are indeed causal (not just correlated), and (b) tests of whether environmental risks exacerbate inherited vulnerability via gene environment interactions. A strong test of the nurture hypothesis would use a twin design to disentangle and 'control' for genetic influences. A fair test of the nurture hypothesis would use a high-risk design to reach children living in the most severely deprived environments, and would use state-of-the art measures of rearing experiences collected during home visits to families.

Broadly, the E-Risk Study has several innovative design features:

1. The E-Risk Study began with 5-year-old children and has access to data about them from birth. Beginning young was important, because it has been shown that very early onset of problem behaviours characterises the most serious cases of adolescent and adult mental disorders. ERisk is now one of only 6 studies of very young twins, worldwide.

2. The E-Risk sample accurately represents the population of families having children in the 1990's in England and Wales. Each family in the study bears a known statistical relation to a segment of the UK population that the sample was drawn to represent.

3. As an innovative genetic study, E-Risk collected direct measures of an exhaustive array of putative environmental risk factors in families, schools, and neighbourhoods.

4. As an innovative environmental study, E-Risk sampled twins. This allows us to disentangle genetic contributions to risk factors, to ascertain the true extent of their environmental effects.

5. The E-Risk Study collected DNA to add measured genes, which are intended to fill in the black box of genetic variance. Knowledge about how measured genes interact with environmental risks will support applied research and development of new diagnostics and treatments.

The combination of these design features in the E-Risk Study was innovative in 1998, and to our knowledge, the study design remains unique in behavioural genetic research.



Contact details

louise.arseneault@kcl.ac.uk or terrie.moffitt@duke.edu


UK Medical Research Council (MRC) 

Page last updated: 12 May 2016


  • Gender: Male, Female
  • Sample size: 0-4,999
  • Age: 0 to 9, 10 to 19
  • Anthropometric: Height, Weight, Waist circumference, Hip circumference, Blood pressure
  • Psychological: Mental health, Cognitive function
  • Lifestyle: Smoking, Physical activity, Dietary habits, Alcohol
  • Socio-economic: Occupation, Finances, Family circumstances, Housing, Education, Ethnic group, Marital status, Social support
  • Biological samples: Blood, Other