The interdependence among eating behaviour (EB), physical activity (PA) and sedentary time (ST) suggests simultaneously identifying homogeneous profiles and describing their changes. This study aimed to (1) identify cross-sectional lifestyle behaviour profiles and their 2-year changes among French school-age adolescents and (2) identify factors associated with these profiles and changes. Longitudinal data from adolescents who participated in the PRomotion de l’ALIMentation et de l’Activité Physique trial were used. PA and ST were assessed with the International Physical Activity Questionnaire and EB with a FFQ. Profiles at baseline and their changes were identified by latent transition analysis. Multinomial logistic regression models were used to identify factors associated with profiles and their changes. Among 2390 adolescents included (14–18 years), five baseline profiles that differed mainly in EB were identified: ‘healthy diet and high PA (7·9 %)’, ‘big eater and moderate to high PA (23·8 %)’, ‘healthy diet and low PA (31·2 %)’, ‘restrictive diet and moderate PA (20·6 %)’ and ‘sugar products, nibbling and moderate PA (16·5 %)’. Young adolescents, those who were overweight or obese and socially advantaged, were more in the ‘healthy diet and low PA’ than others. Boys, older and socially less advantaged adolescents exhibited more ‘unfavourable’ than ‘mixed’ changes, while adolescents with overweight or obesity had less ‘unfavourable’ than ‘mixed’ changes. In conclusion, adolescents were twice the number in the least than the most favourable profile. Findings highlighted the importance of EB among adolescents and suggest taking adolescents’ sociodemographic and weight characteristics into account in interventions aimed at acting on adolescents’ behaviours.

中文翻译：

---

青春期生活方式行为的纵向模式：潜在转变分析

饮食行为 (EB)、体力活动 (PA) 和久坐时间 (ST) 之间的相互依赖性表明同时识别同质分布并描述它们的变化。本研究旨在 (1) 确定法国学龄青少年的横断面生活方式行为概况及其 2 年的变化，以及 (2) 确定与这些概况和变化相关的因素。使用了来自参与 PROmotion de l'ALImentation et de l'Activité Physique 试验的青少年的纵向数据。PA 和 ST 使用国际体育活动问卷进行评估，EB 使用 FFQ 进行评估。通过潜在转换分析确定基线的配置文件及其变化。多项逻辑回归模型用于识别与配置文件及其变化相关的因素。在包括的 2390 名青少年（14-18 岁）中，确定了 5 个主要在 EB 方面不同的基线资料：“健康饮食和高 PA（7·9 %）”、“大食量和中度至高 PA（23·8 %） '、'健康饮食和低 PA (31·2 %)'、'限制性饮食和适度 PA (20·6 %)' 和'糖制品、小食和适度 PA (16·5 %)'。年轻的青少年，那些超重或肥胖且具有社会优势的人，比其他人更倾向于“健康饮食和低 PA”。男孩、年龄较大和社会地位较低的青少年表现出比“混合”变化更“不利”的变化，而超重或肥胖的青少年的“不利”变化比“混合”变化要少。总之，青少年的人数最少是最有利的人数的两倍。