Consumption of added sugars is decreasing in the United States

Abstract

Background: The consumption of added sugars (caloric sweeteners) has been linked to obesity, diabetes, and heart disease. Little is known about recent consumption trends in the United States or how intakes compare with current guidelines.

Objective: We examined trends in intakes of added sugars in the United States over the past decade.

Design: A cross-sectional study of US residents ≥2 y of age (n = 42,316) was conducted by using dietary data from NHANES 1999-2008 (five 2-y cycles) and data for added-sugar contents from the MyPyramid Equivalents Database. Mean intakes of added sugars (grams and percentage of total energy intake) were weighted to obtain national estimates over time across age, sex, and race-ethnic groups. Linear trends were tested by using Wald's F tests.

Results: Between 1999-2000 and 2007-2008, the absolute intake of added sugars decreased from a mean (95% CI) of 100.1 g/d (92.8, 107.3 g/d) to 76.7 g/d (71.6, 81.9 g/d); two-thirds of this decrease, from 37.4 g/d (32.6, 42.1 g/d) to 22.8 g/d (18.4, 27.3 g/d), resulted from decreased soda consumption (P-linear trend <0.001 for both). Energy drinks were the only source of added sugars to increase over the study period (P-linear trend = 0.003), although the peak consumption reached only 0.15 g/d (0.08, 0.22 g/d). The percentage of total energy from added sugars also decreased from 18.1% (16.9%, 19.3%) to 14.6% (13.7%, 15.5%) (P-linear trend <0.001).

Conclusion: Although the consumption of added sugars in the United States decreased between 1999-2000 and 2007-2008, primarily because of a reduction in soda consumption, mean intakes continue to exceed recommended limits.

FIGURE 1.

Mean (±SE) trends in the proportion of total energy intake consumed as added sugars and as sugar-sweetened beverages (SSBs) in US children and young adults by age group, NHANES 1999–2000 to 2007–2008. Results of linear trends by using Wald's F tests were P < 0.0001 for total added-sugar and SSB intakes for all subjects aged <35 y. In subjects aged 35–54 y, P-trend = 0.0005 for total added-sugar intake and P-trend = 0.001 for SSB intake; in subjects aged ≥55 y, P-trend = 0.002 for total added-sugar intake and P-trend = 0.01 for SSB intake.

FIGURE 2.

Mean (±SE) trends in the proportion of total energy intake consumed as added sugars and as sugar-sweetened beverages (SSBs) by race-ethnic group, NHANES 1999–2000 to 2007–2008. The results of linear testing by using Wald's F tests were P < 0.0001 for total added sugar in all races and ethnicities. P-trend for SSB intake was <0.0001 for non-Hispanic blacks and other Hispanics, 0.0006 for Mexican Americans, and 0.0002 for non-Hispanic whites.

FIGURE 3.

Mean (±SE) trends in the proportion of total energy intake consumed as added sugars and as sugar-sweetened beverages (SSBs) by income quartile, NHANES 1999–2000 to 2007–2008. The results of linear trends by using Wald's F tests were P < 0.0001 for total added-sugar intake in all income groups. P-trend for SSB intake was <0.0001 for all but the lowest income group for whom the P-trends for total added sugar and SSB intake were 0.0009 and 0.02, respectively.