Exploring the Role of Sugar-Sweetened Beverage Consumption in Obesity among New Yorkers Using Propensity Score Matching

Marissa Burgermaster*, Hiershenee Bhana, M. Dot Fullwood, Diego A. Luna Bazaldua, Elizabeth Tipton

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

1 Scopus citations


Background Results from clinical trials have shown that sugar-sweetened beverages (SSBs) lead to increased body mass index (BMI) and obesity. This relationship has yet to be explored in observational data for nonclinical populations of adults. Objective To compare adults who drank 4+ SSBs daily to those who drank 0 in the population of adults in New York City, and to better understand adult risk factors associated with higher daily SSB consumption and BMI. Design Secondary analysis of cross-sectional data using propensity score matching. Participants/setting The 2009 New York City Community Health Survey (N=9,934) was used. Main outcome measure BMI. Statistical analyses For each participant who consumed 4+ SSBs daily, propensity score matching identified matched comparisons who did not drink any SSBs. BMI in unadjusted and matched pairs was tested using t tests. A post hoc analysis compared features of those likely to drink SSBs and those not likely to drink SSBs. Results In unmatched analyses, participants who consumed 4+ SSBs daily (n=475) had higher BMI than those who consumed 0 SSBs (n=3,818; BMI difference=1.4±0.29; t value=4.81; P<0.001); however, when compared with similar participants using nearest neighbor with replacement matching (n=1,062), the difference between those who consumed 4+ SSBs daily and those who consumed none decreased (BMI difference=0.37±0.36; t value=1.01; P=0.32). Analyses also indicated that those likely to drink SSBs and those unlikely to drink SSBs differed in several important characteristics, including sex, age, race, ethnicity, socioeconomic status, education, diet, and exercise. Conclusions The data preclude strong causal conclusions about the role of SSB in obesity. However, our results suggest that there is a subset of participants demographically and behaviorally similar with higher BMI regardless of their self-reported SSB intake. In addition to targeting SSBs, public health policies and programs should identify and address other modifiable aspects of this profile and tailor approaches to the groups identified to be most affected by high BMI.

Original languageEnglish (US)
Pages (from-to)753-762
Number of pages10
JournalJournal of the Academy of Nutrition and Dietetics
Issue number5
StatePublished - May 2017


  • Health disparities
  • Obesity
  • Propensity score matching
  • Sugar-sweetened beverages
  • Tailoring

ASJC Scopus subject areas

  • Food Science
  • Nutrition and Dietetics

Fingerprint Dive into the research topics of 'Exploring the Role of Sugar-Sweetened Beverage Consumption in Obesity among New Yorkers Using Propensity Score Matching'. Together they form a unique fingerprint.

Cite this