Table 4

Longitudinal studies

CitationTypeQuality assessmentCountryParticipantsStudy designOutcome measuresMain findings
Choi et al (2016)58Short paperModerateUSABoys and girls (n=894 at T1, n=780 at T2)
16–18 years (mean 17.0 years, SD 0.77)
Longitudinal mixed mode survey (web, telephone and paper) at baseline and 12 monthsAlcohol and ED use, binge drinkingED use was positively associated with alcohol use. After controlling for alcohol use at baseline, the effect size of ED use in the past month decreased or became non-significant. ED use at baseline predicted the number of drinking days, but not past month binge drinking or average drinks per drinking day, at follow-up
Martz et al (2015)60Full paperModerateUSABoys and girls (n=6498)
17–18 years
Monitoring the Futures surveys completed by 12th grade students in 2012 and 2013Use of alcohol mixed with EDs (AmED), academic and social factors, other substance use, unsafe drivingMales had significantly greater odds of any AmED use than females, and White or Hispanic students had significantly greater odds than Black students. AmED use was significantly associated with greater odds of driving violations and accidents after alcohol use, controlling for all other variables
Miyake and Marmorstein (2015)53Full paperModerateUSABoys and girls (n=144 at T1, n=127 at T2)
11–13 years
Classroom-based survey at baseline and 16 monthsUse of caffeinated drinks and alcohol, sensation-seeking, parental monitoringFrequency of ED consumption at baseline predicted increases in frequency of alcohol consumption 16 months later. Lower levels of parental monitoring were associated with higher levels of ED consumption and later frequency of alcohol use
There were significant associations between baseline levels of sensation-seeking and frequency of ED consumption, but not later alcohol use
Richards and Smith (2016)46Full paperStrongUKBoys and girls (n=2610 at T1, n=2307 at T2)
11–16 years
Longitudinal study of diet in secondary school children, using the Diet and Behaviour Scale, at baseline and 6 monthsDietary intake of common foods and drinks, exercise frequency, and self-assessed mental health (at T2 only)Drinking EDs once a week or more was significantly associated with being male, older, having special educational needs and being eligible for free school meals. Those who consumed EDs once a week or more slept for fewer hours per night, achieved lower school attendance, higher Junk Food scores, and exercised more frequently (though the latter effect was only marginally significant)
ED consumption alone was not predictive of stress, anxiety or depression at 6-month follow-up. However, in the regression analyses, high stress levels were associated with being a member of the frequent EDs/infrequent breakfast condition