Plain water consumption is associated with lower intake of caloric beverage: cross-sectional study in Mexican adults with low socioeconomic status

Daniel Illescas-Zarate, Juan Espinosa-Montero, Mario Flores, Simon Barquera

Plain water (PW) should be the main beverage consumed by the population. However, consumption of caloric beverages (CB) has increased considerably worldwide. The purpose of this paper is to analyze the association between CB and PW intake in Mexican adults with a low socioeconomic status (SES). In a cross-sectional design, beverage consumption was evaluated with a 24-h beverages recall using the five-step multiple-pass method recommended by the U.S. Department of Agriculture. Physical activity, anthropometric and sociodemographic information were obtained. CB was defined as those beverages that provide energy, with the exception of low-fat milk and beverages with noncaloric sweeteners. Participants were classified into five groups according to their PW consumption (nondrinkers and four quartiles). Differences between groups were evaluated with ANOVA and Bonferroni tests for multiple comparisons among quartiles. A two-stage Heckman regression model was designed with robust standard errors, adjusting for potential confounders. A total of 1108 adults between 21 and 59 years of age were evaluated. A negative association was noted between PW intake and CB consumption (p <0.001) with the exception of natural juice, which was positive (p <0.01) and sodas that no differences were found between quartiles. Specifically, for every milliliter of PW, the intake of CB was 3.4, 1.3, 0.68 and 0.38 mL in each quartile, respectively (p <0.001). In Heckman's model, PW consumers were 0.5 times less likely to consume CB (p = 0.029). This probability increased to 0.9 for low-fat milk, skim milk and beverages without added sugar (LFM-BWAS) consumers (p <0.001). Also, for every 100 mL of PW consumption, CB intake diminished by 20 mL (p <0.001). In turn, for every 100 mL of LFM-BWAS consumption, a reduction of 47 mL in CB was observed (p <0.001). Higher PW consumption was associated with lower CB consumption. This association suggests that future studies are warranted to determine if increasing PW intake in a low SES Mexican population can reduce intake of CB.