Associations of food group intakes with serum carbon isotope ratio values in youth: results from two prospective pediatric cohort studies.

Abstract

Background: The carbon isotope ratio (CIR) is a candidate biomarker for sugar sweetened beverage (SSB) intake in the U.S. However, research specific to youth, who differ in their physiology and dietary patterns compared to adults, is lacking.

Objectives: We evaluated longitudinal associations of SSB intakes across childhood/adolescence with serum CIR. We also explored the relationship between other dietary intakes and serum CIR.

Methods: Data were from participants in two longitudinal, pediatric cohorts in Colorado: Exploring Perinatal Outcomes among CHildren Study (EPOCH) (visits at median 10 and 16 y, n=150) and Healthy Start Study (visits at median 5 and 9 y, n=166). Serum CIR was measured using isotope ratio mass spectrometry. Diet was assessed by food frequency questionnaires (EPOCH) or 24-hour diet recalls (Healthy Start). We assessed associations of longitudinal dietary intakes (log2-transformed, standardized) with serum CIR using linear mixed models adjusted for age, sex, and energy intake, and associations of change values between visits using linear regression models.

Results: In linear mixed models, higher SSB intake across visits was associated with higher serum CIR in both cohorts (β [95% CI]: 0.11[0.06,0.15] in EPOCH and 0.14[0.07,0.21] in Healthy Start). Higher meat intake and a higher dietary animal protein ratio were also positively associated with serum CIR in both cohorts (β [95% CI]: 0.08[0.05,0.12] and 0.18[0.13,0.23] in EPOCH; 0.08[0.01,0.16] and 0.28[0.21,0.35] in Healthy Start). In change analyses, there were positive associations for changes in the dietary animal protein ratio between visits with changes in serum CIR in both cohorts, but not for changes in SSB intake.

Conclusions: Our findings support serum CIR as a potential biomarker of SSB intake in youth cross-sectionally; however, there was not a strong link between change values over longer-term follow-up. Meat/animal protein intake was also consistently and, at times, more strongly associated with serum CIR.