Changes in alcohol consumption among Belgian adults participating in the internet-based one-month-abstinence campaign...

Temporary abstinence campaigns (TAC) are a promising approach to reduce alcohol consumption. This study examined changes in (at-risk) alcohol consumption and determinants of alcohol consumption among Belgian adults participating in the Tournée Minérale Campaign (TMC, Belgian TAC). In addition, moderating effects of age, sex, education level, successfully refraining from alcohol during TMC, risk drinking, drinking levels and binge drinking were investigated.

TMC is a fully automated mass media prevention campaign challenging Belgian adults to refrain from alcohol during the month February. This study includes a baseline, post- (four weeks after TMC) and follow-up (six months after TMC) test using online recruitment and an online questionnaire. Participants were self-selected by signing up on the TMC website. Repeated measures ANCOVAs were used to examine changes over time in alcohol consumption (log transformed due to skewness) and determinants among TMC participants (n = 8,730, 48.4 ± 12.9 years old, 37.4% males) who filled in all three measurements. Moderation effects were assessed for age, sex, education level, successfully refraining from alcohol during TMC, risk drinking, drinking levels and binge drinking at baseline. McNemar tests were used to examine the change in prevalence of risk drinking and binge drinking.

TMC participants showed a significant decrease in weekly alcohol consumption (mean exp(x)-1) and CI) from baseline (6.2 [6.4,6.7]) to post (4.2 [4.1,4.3], F = 22.0, p < .001 d = -2.4 [-2.4,-2.3]) and from baseline to follow-up (5.1 [4.9,5.2], F = 24.2, p < .001, d = -1.5 [-1.5,-1.4]), and an increase from post to follow-up (F = 21.1, p < .001). In the short term (from baseline to post), the decreases were stronger among older, male and lower educated TMC participants, those successfully refraining during TMC, higher risk drinkers and binge drinkers. In the medium term (from baseline to follow-up), changes in alcohol consumption were moderated by education (stronger among lower educated) and risk drinking (increase in lower risk drinkers and decrease in higher risk drinkers). Participants' baseline drinking level moderated changes in alcohol consumption. Both short and medium term increases in alcohol consumption were observed among those with a low drinking level (those who drank 0 to < 4 glasses at baseline), while short and medium term decreases were observed among those with higher drinking levels, i.e. those with moderate (≥ 4 to ≤ 10 glasses), high (> 10 to ≤ 17 glasses) and highest (> 17 glasses) drinking level at baseline. A reduction in the proportion of risk drinkers and binge drinkers and changes in determinants of alcohol consumption (e.g. attitudes towards drinking less alcohol and habit of drinking alcohol) were observed among TMC participants.

TMC participants reported favourable changes in (the underlying determinants of) alcohol consumption in the short and medium term. Higher level drinkers appear to be especially attracted to participate in TMC and to experience the strongest decreases in alcohol consumption. However, without comparison with a control group of non-participants or comparison points, we cannot simply attribute the changes over time to participation in TMC. Future research with a stronger study design is needed to rule out possible bias (e.g., selection bias, seasonal effects or regression to the mean) and to examine mechanisms and longer term effects of TACs.