Dose-Response and Substitution Analyzes of Sweet Beverage Consumption and Body Weight in Dutch Adults: The Lifelines Cohort Study

BACKGROUND/METHODS: Prospective studies investigating sweet beverages and body weight associations show inconsistent results. Within the SWEET project, we examined prospective dose-response associations of sugar-sweetened beverages (SSB), low/no-calorie beverages (LNCB), and fruit juice with body weight-related outcomes among 78,286 Dutch adults followed for ~4 years. Baseline intakes were assessed using a validated food-frequency questionnaire (FFQ) with 150 ml representing a standard serving. Outcome variables were body weight change, waist circumference change, overweight/obesity, and abdominal obesity. Associations were investigated by using linear and non-linear dose-response analysis, as well as substitution models while adjusting for multiple socio-demographic, lifestyle, health, and dietary variables. RESULTS: Participants were 46 ± 13 (mean ± SD) years old and 60% were women. Adjusted dose-response analyzes indicated an association between SSB and LNCB, and both body weight (+0.02 kg/year; SE 0.01 and +0.06 kg/year; SE 0.01) and waist circumference changes (+0.04 cm/year; SE: 0.01 and +0.11 cm/year; SE: 0.01). Associations for overweight/obesity and abdominal obesity incidence were +3% (95%CI: 1.00–1.06) and +2% (95%CI: 0.99–1.06) for SSB and +8% (95%CI: 1.06–1.11) and +5% (95%CI: 1.03–1.07) for LNCB, respectively. Substitution of SSB with LNCB was associated with higher weight change (+0.04 kg/year), waist circumference change (+0.09 cm/year), overweight/obesity incidence (+6%), but not abdominal obesity incidence. For fruit juice, we observed beneficial associations for intake levels below ~1 serving/day with weight, waist circumference change, and overweight/obesity incidence, and no association with abdominal obesity. Subsequent substitution analyzes indicated a small beneficial association for the replacement of SSB with fruit juice on weight (−0.04 kg/year) and waist circumference (−0.04 cm/year), but not with other outcomes. CONCLUSIONS: Overall, our results suggest that habitual consumption of both SSB and LNCB may adversely affect weight-related outcomes. In contrast, fruit juice consumption <150 ml may be beneficial with respect to weight and waist circumference.