Lifestyle assessment

Lifestyle was considered a latent variable, that is, not directly observable, and was evaluated by latent class analysis (LCA)⁽¹⁷⁾. With the information from the manifested variables, we elaborated on a statistical model that allowed estimating the probability of a given individual belonging to each of the latent variable categories⁽¹⁸⁾.

In this study, the lifestyle LCA variables were PA, sedentary behaviour, number of meals, alcohol and tobacco consumption. All these variables were evaluated during eight consecutive days. The first day of evaluation was discarded to minimise the Hawthorne effect, which consists of changing behaviour to fulfil the expectations of the study⁽¹⁹⁾.

The PA was evaluated by the Digiwalker SW 200 pedometer, using a cut-off value of 11 700 to determine if the number of steps could be considered an active or inactive behaviour^(19,20). The 24-h recall complemented this evaluation⁽²¹⁾. The pedometer recorded participant’s scored activities performed in a 24-h period (every 15 min); moderate-to-vigorous PA (MVPA) were defined as those with a metabolic equivalent equal to or above 3. The metabolic equivalent corresponds to the metabolic-rate multiple needed for an individual to remain at rest. For this study, the adequate average daily time for MVPA considered was at least 60 min⁽²²⁾.

The sedentary behaviour was assessed by ST, cell phone screen time (CT) and sitting time during weekdays and weekends. ST and CT were measured according to the questionnaire proposed by Miranda et al.⁽⁸⁾ which evaluates the time spent per day in front of television, computer, video game and tablets. CT was analysed separately from the other electronic devices; however, both analysis classified the activities as high when the mean time in the evaluated days was ≥120 min/d, which is the standard time defined by the American Academy of Pediatrics⁽²³⁾.

We analysed the sitting time during weekdays and weekends according to section four of the International Physical Activity Questionnaire⁽²⁴⁾. The weighted average of both data allowed us to estimate the sitting time of both weekdays and weekends. The 75th percentile (75th P) was used as the reference value for sitting time classification due to the lack of a specific cut-off point. The 75th P for all days assessed was 585 min.

The number of daily meals was recorded based on breakfast, collation, lunch, afternoon snack, dinner (or snack) and supper. The mean value during the 7 d was calculated and later categorised by the 50th percentile (50th P = 4·0). Values lower than 25thpercentile were considered small number of meals.

Alcohol and tobacco consumption were observed by two short modules of the Global School – Based Student Health Survey⁽²⁵⁾. The answer option represented by the letter ‘a’ for all questions showed that the teenager had never had any type of alcohol and tobacco use. The other responses were coded with a numerical score of increasing order to be able to quantify the consumption of alcoholic beverages and exposure to tobacco.