Objective: To identify demographic and consumer characteristics associated with refilling a soft drink at fast-food restaurants and the estimated energy content and volume of those refills.
Design: Logistic and linear regression with cross-sectional survey data.
Setting: Data include fast-food restaurant receipts and consumer surveys collected from restaurants in New York City (all boroughs except Staten Island), and Newark and Jersey City, New Jersey, during 2013 and 2014.
Subjects: Fast-food restaurant customers (n 11795) from ninety-eight restaurants.
Results: Thirty per cent of fast-food customers ordered a refillable soft drink. Nine per cent of fast-food customers with a refillable soft drink reported refilling their beverage (3 % of entire sample). Odds of having a beverage refill were higher among respondents with a refillable soft drink at restaurants with a self-serve refill kiosk (adjusted OR (aOR)=7·37, P<0·001) or who ate in the restaurant (aOR=4·45, P<0·001). KFC (aOR=2·18, P<0·001) and Wendy's (aOR=0·41, P<0·001) customers had higher and lower odds, respectively, of obtaining a refill, compared with Burger King customers. Respondents from New Jersey (aOR=1·47, P<0·001) also had higher odds of refilling their beverage than New York City customers. Customers who got a refill obtained on average 29 more 'beverage ounces' (858 ml) and 250 more 'beverage calories' (1046 kJ) than customers who did not get a refill.
Conclusions: Refilling a beverage was associated with having obtained more beverage calories and beverage ounces. Environmental cues, such as the placement and availability of self-serve beverage refills, may influence consumer beverage choice.
Keywords: Fast food; Nutrition; Obesity; Sugar-sweetened beverages.