BACKGROUND AND AIMS: Restricting alcohol advertising may reduce alcohol consumption and related ill-health. However, advertisements for alcohol-free and low-alcohol drinks (NoLos) with brand-matched alcoholic versions are typically exempt from restrictions, which could lead to surrogate marketing (adverts for NoLo beverages also promoting brand-matched alcoholic options). This study measured the impact of advertisements for brand-matched NoLo beverages on product selections in a simulated online supermarket, in the UK.

DESIGN, SETTING, PARTICIPANTS AND INTERVENTION: We conducted a randomised controlled trial with 1638 UK regular alcohol consumers (aged 18-91), assigned to one of three groups: (1) alcohol adverts (n = 469), where participants viewed an advertisement (embedded within a video) for one of four alcoholic beverages; (2) NoLo adverts (n = 472), for one of four brand-matched NoLo beverages; and (3) unrelated adverts (n = 697), for batteries. Participants then selected food and drinks for a barbecue in a simulated online supermarket.

MEASUREMENTS: We collected data on products selected in the simulated online supermarket (e.g. product name, category, quantity, energy). Risk of alcohol dependence was measured using the Alcohol Use Disorder Identification Test Consumption Questions (AUDIT-C). The primary outcome was selection (yes/no) of advertised alcoholic and NoLo products. Secondary outcomes included the selection (yes/no) of any alcoholic or NoLo products. We also examined the recall of advertised brands and products through survey questions.

FINDINGS: Viewing NoLo, but not alcohol, advertisements statistically significantly increased the odds of selecting the advertised products, compared with unrelated advertisements [alcohol group: odds ratio (OR) = 1.62, 95% confidence interval (CI) = 1.03-2.53, Benjamini-Hochberg-adjusted P = 0.11; NoLo group: OR = 2.18, 95% CI = 1.24-3.91, Benjamini-Hochberg-adjusted P = 0.022]. Although the odds of selecting the alcoholic version of the advertised product were higher in those exposed to NoLo adverts vs. control (OR = 1.48, 95% CI = 0.94-2.33, Benjamini-Hochberg-adjusted P = 0.13), this association did not reach statistical significance. Fifty-nine percent of participants in all three groups recalled the advertised brand. Among those who remembered the brand of advertisement, 96% in the alcohol group also correctly recalled the advertised product, while 44% in the NoLo group reported seeing an advert for the alcoholic version of the product (X = 297.16, P < 0.001, df = 2).

CONCLUSIONS: Exposure to advertisements for brand-matched alcohol-free and low-alcohol drinks increases brand (over product) recall, but, while the direction of effects is consistent with these advertisements promoting the selection of alcoholic beverages, evidence of the impact on alcohol selection is inconclusive.