Rationale The UK Soft Drinks Industry Levy (SDIL) was introduced in response to evidence on the role of sugary drinks in obesity, diabetes and tooth decay. The levy is two-tiered: £0.24/L for drinks containing >8 g/100 ml of added sugar and £0.18/L for drinks containing 5–8 g sugar/100 ml, and directed at manufacturers and importers of soft drinks, in order to encourage reformulation. The levy came into effect in April 2018, but was announced two years before to allow industry time to adapt. The announcement may represent a public health intervention itself.

Aim To examine change in the volume of, and amount of sugar from, household purchases of soft drinks from two years before announcement of the SDIL to two years after.

Methods An interrupted time series design was used. Kantar WorldPanel data from ∼43,000 UK households recorded all food and drink purchases brought home, covering April 2014 to March 2018 (208 weeks). Outcomes were purchased volume of, and sugar from, liable soft drinks in each levy tier as well as non-liable soft drinks and confectionery (a potential high sugar substitute category) per household per week, modelled against the counterfactual of no announcement. Household purchases of toiletries (shampoo, conditioner and liquid soap) were included as a comparator. Regression analyses were weighted to reflect UK purchasing patterns.

Results Immediately after the announcement there was an increase in the volume of (73 ml per household per week, 95%CI: 29,119), and amount of sugar from (11.0 g per household per week, 95%CI: 3.7, 18.4), higher tier drinks and a decrease in the amount of sugar from confectionery (-24.6 g per household per week, 95% CI: -47.1,-2.15). In the period following the announcement there was a significant downward trend in the volume of (0.01 ml per household per week², 95%CI: -0.02,-0.01), and amount of sugar (-0.0008 g per household per week², 95%CI: -0.0012, -0.0004) from, drinks in the lower tier. During the same period sugar from non-liable soft drinks increased (0.0003 g per household per week², 95%CI: 0.00001, 0.00064)

Conclusion The announcement of the SDIL was associated with sustained reductions in the volume of, and amount of sugar from, drinks in the lower levy tier purchased by UK households; and a sustained increase in the amount of sugar from non-liable soft drinks. This likely reflects reformulation by manufacturers so that many drinks previously in the lower tier are now non-liable but still contain sugar. There was no evidence of substitution to confectionary.