Altering the Availability of Healthier vs. Less Healthy Items in Vending Machines

Background: While there is some evidence that increasing the range of healthier foods and drinks and/or decreasing the range of less healthy options may increase healthier choices, more work is needed to establish the reproducibility of any effect. The current study aims to investigate the impact of altering the availability of healthier and less healthy foods and cold beverages in hospital vending machines.

Methods: An adapted multiple treatment reversal design will be used, in which all standard vending machines serving snack foods and/or cold beverages in one hospital in England change the number of slots containing (i) less healthy items and (ii) healthier items over eight 4-week periods. Changes will take place in a two-step process whereby decreases are implemented in a separate study period prior to increases in the contrasting food group. Following a 4-week baseline period, all vending machines will be standardised to have 75% healthier drinks and/or 25% healthier snacks (study period 1). Vending machines (n=9) will be randomly allocated to the order in which they: (1) decrease less healthy foods and increase healthier foods or (2) decrease healthier foods and increase less healthy foods (study periods 2&3 and 5&6). After each decrease-increase pair, machines will return to the standardised 75% healthier drinks and 25% healthier snacks (study periods 4 and 7). Sales data will be obtained via records of machine restocking.

Planned Analysis: The impact of the availability intervention will be assessed in separate linear mixed models for cold drinks and snacks, examining the impact on total energy (kcal) purchased, per restocking interval, with random effects for vending machine.