Assessing the Role of Price Promotion in the Association between Income and Sugar- Sweetened Beverage Purchases

Abstract

Background: In the U.S., there are income disparities in SSB consumption where lower incomes consume more SSBs. Evidence is limited on whether controlling for price promotions and shopping behaviors would help explain / attenuate the relationship between income and SSB consumption. Alternately, it is possible that SSB price promotions serves as an effect modifier and widens income disparities in SSB consumption due to lower income price sensitivity. The objective of this study is to assess whether the income gradient in SSB purchases is explained by the exposure to price promotion and/or different price promotion-related shopping behaviors, and assess the presence of an interaction between price promotion and income on the association with SSB purchases, while controlling for shopping behaviors. Methods: We conducted a cross-sectional study using a large nationally representative sample from the NielsenIQ Consumer Panel (analytic sample N=11,299) in 2016. Annual promotion frequency (defined as percent of the time experiencing SSBs discounts) for each household were derived using weekly store prices in the NielsenIQ Retailer Scanner Database. A stockpiling index was generated for each household to describe bulk shopping behavior. We compared the association between household income per capita and annual per capita SSB purchase with and without adjusting for price promotion and/or the stockpiling index to determine whether the income gradient in SSB purchase can be explained by the exposure to price promotion, or promotion-related shopping behaviors. Results: Nearly 40% of SSBs were purchased with a price discount. More frequent price promotions were linked to higher SSB purchase. Overall, lower income households purchased substantially more SSBs (i.e., household income per capita ≤$15,000, $15,000 to $30,000, $30,000 to $50,000 categories purchased 54%, 22% and 10% more SSBs than household income per capita ≥$50,000, respectively). However, the association between income and purchase of SSBs was strong regardless of adjustment for price promotion and there was no evidence that the association varied by exposure to price promotions. SSB shopping behaviors were different by income. Lower income households made more frequent trips to purchase SSBs and smaller purchases (i.e., made 12 trips per year, purchased 58 oz. per capita on each trip) while higher income households the opposite (i.e., made 5 trips per year, purchased 72 oz. per capita on each trip). After adjusting for covariates and the stockpiling index, the income gradient in SSB purchase was slightly attenuated but still significant (i.e., households in income per capita ≤$15,000, $15,000 to $30,000, $30,000 to $50,000 categories purchased 32%, 17% and 4% more SSBs than households with income per capita ≥$50,000, respectively). Conclusions: Income disparities in SSB consumption remained after covariate adjustment and after accounting for exposure to price promotion and different shopping behaviors.