Alcohol & Alcoholism Vol. 43, No. 6, pp. 641–646, 2008 doi: 10.1093/alcalc/agn077 Advance Access publication 8 October 2008 COGNITIVE AND BEHAVIOURAL EFFECTS The Self-Rating of the Effects of Alcohol Questionnaire as a Predictor of Alcohol-Related Outcomes in 12-Year-Old Subjects Marc A. Schuckit 1,∗ , Tom L. Smith 2 , Ryan S. Trim 3 , Jon Heron 4 , Jeremy Horwood 5 , John Davis 6 , Joseph Hibbeln 7 and the ALSPAC Study Team 8 1 VA San Diego Healthcare System/University of California, San Diego, 2 University of California, San Diego, 3 VA San Diego Healthcare System, 4 Avon Longitudinal Study of Parents and Children (ALSPAC), University of Bristol, UK, 5 University of Bristol, UK, 6 University of Illinois at Chicago, 7 National Institute of Alcohol Abuse and Addiction (NIAAA), NIH Bethesda, Maryland, and 8 University of Bristol, UK ∗ Corresponding author: Department of Psychiatry, University of California, 3350 La Jolla Village Drive, San Diego, CA 92161-2002, USA. Tel: +1-858-552-8585 extn. 7978; Fax: +1-858-552-7424; E-mail: mschuckit@ucsd.edu (Received 25 January 2008; first review notified 8 July 2008; in revised form 1 August 2008; accepted 25 August 2008; advance access publication 8 October 2008) Abstract — Aims: A low level of response (LR), or low sensitivity, to alcohol as established by alcohol challenges has been shown to predict future heavier drinking, alcohol-related problems and alcohol use disorders. To date, only one study has evaluated the predictive validity of a second measure of LR as determined by the Self-Report of the Effects of Alcohol (SRE) Questionnaire. The current analyses evaluate the ability of SRE scores as determined at age 12 to predict heavier drinking and alcohol-related problems 2 years later in a sample from the United Kingdom. Methods: The subjects were 156 boys (54.5%) and girls from the Avon Longitudinal Study of Parents and Children (ALSPAC) who had reported consuming one or more standard drinks by age 12 and who were followed up 2 years later. Results: The age 12 SRE scores correlated with the number of drinks per week, maximum drinks and the number of alcohol problems both at baseline and at age 14 follow-ups. In these evaluations, a larger number of drinks required for effects on the SRE (i.e. a lower LR per drink consumed) related to heavier intake and alcohol-related difficulties. Simultaneous entry multiple regression analyses revealed that the age 12 SRE score maintained a significant relationship with age 14 higher number of drinks per week and the number of alcohol problems even when the age 12 values for alcohol intake and problems were used as covariates. Conclusion: The SRE scores appear to have value in predicting future heavier drinking and alcohol problems in 12-year olds that go beyond the information offered by the earlier drinking pattern alone. INTRODUCTION A person’s level of response (LR) to alcohol is one of several genetically influenced characteristics that predict future heav- ier drinking and alcohol problems (Heath et al., 1999; Schuckit and Smith, 2000; Schuckit, 2002). The hypothesis that a low LR to alcohol would relate to higher intake of this drug and a greater risk for problems was supported by animal studies (Baldwin et al., 1991; Barr et al., 2003) and by most human investigations (Pollock, 1992; Ehlers et al., 1999; Erblich and Earleywine, 1999; Wall et al., 1999). The initial human studies of LR used alcohol challenges where the response was deter- mined by observing the degree of alcohol-related changes in a range of characteristics after consuming two to four drinks (Schuckit and Gold, 1988). Most alcohol challenge-based eval- uations have indicated that a low LR to alcohol correlated with family histories of alcohol use disorders (AUDs) and related to future heavier drinking in the subjects, even after control- ling for the past drinking pattern (Pollock, 1992; Rodriguez et al., 1993;Volavka et al., 1996; Erblich and Earleywine, 1999; Heath et al., 1999; Schuckit and Smith, 2000; Eng et al., 2005; King et al., 2006). Both human and animal studies using alcohol challenges have supported the importance of genes as contrib- utors to the LR to alcohol, and in alcohol-naive animals, a low LR was associated with higher alcohol consumption (Baldwin et al., 1991; Schuckit et al., 2001; Schuckit, 2002; Barr et al., 2003). However, the evaluation of LR using alcohol challenges is time consuming, expensive and limited to healthy subjects old enough to give informed consent, with most work focusing on 18- to 25-year-old individuals (Wall et al., 1999; Schuckit, 2002). These limitations contributed to the development of a sim- pler retrospective self-report measure of LR by recording the number of drinks a person reports as having been required for each of four possible effects of alcohol early in the drinking career (Schuckit et al., 1997, 2001, 2006; Bernard et al., 2007). In this work, a standard drink is defined as the amount of an alcoholic beverage that contains 10–12 g of ethanol, and the drinking response early in life is arbitrarily defined as the ap- proximate first five times or so of consuming at least one full drink. The LR scores from the Self-Report of the Effects of Al- cohol (SRE) Questionnaire have correlated as high as 0.6 with alcohol challenges, have a 1-year retest reliability of ∼0.80, are familial and potentially genetically influenced, and have correlated with heavier drinking and alcohol problems even in subjects as young as age 12 and for whom alcohol tolerance was not likely to have developed (Schuckit et al., 1997, 2001, 2005a, 2005c, 2006). SRE-based LR scores have been useful in analyses searching for genes related to LR and performed similarly to alcohol challenge LR values in structural equation models evaluating how LR related to additional life domains in relating to heavier drinking and alcohol-related problems (Schuckit et al., 2001, 2005b; Hinckers et al., 2006). In these analyses, a larger number of drinks required for effects on the SRE are the equivalent of a low LR at a given blood alcohol concentration on alcohol challenges. While the ability of alcohol challenge-based low LR scores to predict heavier drinking and more alcohol problems is sup- ported by all studies to date (Schuckit, 2002), less is known about how well scores on the SRE predict these outcomes. One study presented the results from a 5-year follow-up of 95 18- to 35-year-old subjects (mean age 25) from the US-based C  The Author 2008. Published by Oxford University Press on behalf of the Medical Council on Alcohol. All rights reserved