Changes in Sugar-sweetened Beverage Intake and Metabolic Health: Improving Metabolic Profile Very Effortlessly

Sugar-sweetened beverages (SSBs) have been epidemiologically linked to serious health problems including heart disease, liver disease, and diabetes. This study will recruit frequent SSB drinkers who are employees at the University of California, San Francisco (UCSF) and measure markers of metabolic health on two occasions, 10 months apart. During this time, UCSF will cease selling SSBs at all campus and medical center locations. Additionally, at the first assessment half of participants will be randomly assigned to a brief intervention to help reduce SSB consumption, consisting of a 10-20 minute semi-structured interview designed to share health information about SSBs, elicit motivations to reduce consumption, and help set concrete plans to reduce consumption. We will be able to compare changes in metabolic health among those who do and do not reduce SSB consumption. This will make a unique contribution to the growing evidence regarding both the effects of SSB consumption on health and the modifiability of SSB-related health conditions. Two key clarifications about the design of this study: 1. The cessation of sugar-sweetened beverage sales is a pre-existing workplace initiative. This study is taking advantage of its initiation but it is not part of the study intervention. 2. The primary predictor in this study is SSB consumption, not condition assignment. Although we will report on the efficacy of the brief intervention, the main goal of the study is to examine the correlational relationship between changes in SSB consumption and changes in health, whether the reduction is associated with the intervention or with other factors. Planned analyses: 1. For all outcomes other than sugar consumption and SSB consumption (see below), the primary analytic strategy will be to correlate changes in SSB consumption with changes in the outcome. 2. We will additionally assess the efficacy of the university initiative and the brief counseling intervention using a linear mixed regression model estimating the effects of 1) time (pre to post), representing the effect of the university SSB initiative, and 2) assignment to the brief intervention interacting with time. For both types of analysis, we will include demographic and job-related covariates based on fit criteria to guide model selection.