Abstract Body: Introduction
Ultra-processed food (UPF) intake is high among young populations. This is the first study characterizing UPF intake and assessing its contribution to the traditional diet of Costa Rican adolescents.

Hypothesis
UPF intake varies across Costa Rican adolescents' characteristics, mealtimes, and days of the week, and higher UPF intake is inversely associated with adherence to the traditional Costa Rican diet.

Methods
Using cross-sectional data from 3-day food records from 818 adolescents (13-18y) from San José, Costa Rica, we calculated the daily % of total energy intake (TEI) from Nova-defined UPF, the Traditional Costa Rica Adolescents Diet Score (TCRAD; higher score reflecting higher adherence), nutrients, and fiber. The multivariable-adjusted mean UPF intakes were compared across participants’ characteristics, mealtime, and days of the week, and assessed for association with TCRAD (continuous score and by adherence categories) and dietary components.

Results
UPF intake represented 35.9% of the TEI and was higher (p<0.05) on weekdays (vs. weekends), girls (vs. boys), urban areas (vs. rural), and higher socioeconomic status. UPF were consumed mostly as snacks (43.5%), in the morning (58.5%) and afternoon (41.5%). Each 10-percentage-point increase in the UPF energy contribution was associated with -0.44 (SE=0.04; p<0.0001) lower TCRAD score. The adjusted mean (SE) legume, fruit, vegetable, and vegetable oil intakes (g/1000 kcal/d) associated (p<0.0001) with a 10-percentage-point increase in UPF intake, respectively, were: -5.73 (0.53), -7.99 (1.39), -4.21 (0.89), and -0.58 (0.04). Positive associations (p<0.0001) were observed for UPF intake and the consumption (g/1000 kcal/d) of sugar-sweetened beverages [42.2 (3.0)], fast-food [6.3 (0.7)], desserts/pastries [3.6 (0.5)], bread/cookies [3.0 (0.5)], and snacks [1.5 (0.2)]. UPF were inversely associated with fiber [-0.12g/1000 kcal/d (0.04); p=0.003] and beneficial micronutrients (p<0.05), and directly associated (p<0.0001) with added sugars [6.0 g/1000 kcal/d (0.4)] and trans fats [0.07 g/1000 kcal/d (0.01)]. The multivariable-adjusted UPF intake was lower at moderate (38.5%, p<0.0001) and high (29.9%, p<0.0001) adherence to the TCRAD, compared with at low adherence (43.9%).

Conclusion
Our results support widespread trends of UPF displacing healthy diets. Reducing UPF consumption, particularly during school hours, and promoting healthy traditional diets are crucial for Costa Rican adolescents.