Clustering of diet, physical activity and sedentary behavior among Brazilian adolescents in the national school - based health survey (PeNSE 2015)

Matias et al. BMC Public Health (2018) 18:1283 https://doi.org/10.1186/s12889-018-6203-1 RESEARCH ARTICLE Open Access Clustering of diet, physical activity and sedentary behavior among Brazilian adolescents in the national school - based health survey (PeNSE 2015) Thiago Sousa Matias1*, Kelly Samara Silva1, Jaqueline Aragoni da Silva1, Gabrielli Thais de Mello1 and Jo Salmon2 Abstract Background: There is a lack of evidence regarding clusters of health-related behaviors among adolescents from low, lower-middle, and upper-middle income countries. This study aimed to identify clustering patterns of healthrelated behaviors (diet, physical activity [PA] and sedentary behavior [SB]) and association with sociodemographic variables among a population-based sample of Brazilian adolescents. Methods: Cross-sectional data from the 2015 National School-Based Health Survey (PeNSE). A total of 102,072 (females: 51.7%) students in ninth-grade (age: 14.3 ± 1.1 years-old) enrolled in public and private schools were investigated in this study. Healthy and unhealthy diet, PA and SB were measured using a validated questionnaire. Two-step cluster analysis was conducted to identify lifestyle patterns. The methodology for complex analysis and weighting was used to inferential statistical procedures. Multinomial logistic regression assessed associations between sociodemographic factors and the clusters. Results: Three reliable and meaningful clusters were identified and labelled as follows: (1) health-promoting SB and diet (32.6%); (2) health-promoting PA and diet (44.9%), and (3) health-risk (22.5%). Compared to boys, girls were less likely to be in clusters 1 (OR = 0.85; 95% CI = 0.78–0.93, p < 0.001) and 2 (OR = 0.43; 95% CI = 0.40–0.46, p < 0.001) than the health-risk cluster. Higher socioeconomic status was positively associated with health-promoting PA and diet, and negatively related to health-promoting SB and diet. Older adolescents were more likely to be in cluster 1 than in cluster 3, compared to younger adolescents. Conclusion: Approximately one-quarter of the population (health-risk cluster) reported engaging in multiple risk behaviors. Interventions may need to be tailored to specific adolescent groups, especially considering sociodemographic differences. Keywords: Cluster analysis, Diet, Exercise, Sedentary lifestyle, Adolescent Background Low levels of physical activity (PA), high levels of sedentary behavior (SB) and poor dietary habits are important contributors to several adolescent health problems, such as obesity and cardiovascular risk factors [1]. However, these behaviors do not occur in isolation and there is often synergy between them [2, 3]. An important * Correspondence: 1 Department of Physical Education, School of Sports, Federal University of Santa Catarina, Florianópolis, Brazil Full list of author information is available at the end of the article issue that needs more attention is to investigate the combined occurrence of these behaviors [4, 5]. Indeed, a combination of behaviors will be able to better predict an individual’s overall healthy lifestyle [6–8]. Research on clustering of health behaviors has increased in recent times [9, 10] since it helps to deepen the understanding of how to promote health, allowing a more integrated approach [11]. Cluster analysis is a potential tool for organizing individuals into mutually exclusive groups by taking into account similarities in characteristics and behaviors [5, 9]. In this © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Matias et al. BMC Public Health (2018) 18:1283 way, it is possible to verify which behaviors coexist among individuals [4]. A previous review showed that diet, PA, and SB tend to cluster in adolescents in a complex way, resulting in both healthy and unhealthy groups [5]. Also, observational studies have explored clustering patterns and thus observed a co-occurrence of positive and negative health-related behavior [6, 12–14]. For instance, in a study from 10 European countries of 2084 adolescents, the authors identified mixed clusters of unhealthy and healthy behavioral patterns (e.g. The cluster 3 was observed to be comprised by active adolescents, however, with low diet quality and high sedentary level; contrarily, cluster 4 was composed by inactive adolescents, however, with high diet quality and low sedentary levels) [4]. A study with 7372 children aged 9–11 years put out a concern due to a notable commonality in health-related behavior patterns was found across the 12 countries: all of them presented a cluster characterized by high levels of SB [15]. Despite this evidence, a recent systematic review called attention to the fact that most research has involved the examination of PA and SB, but not dietary factors [5]. That review reported 18 studies involving these behaviors, but only two included PA and diet and none included SB and diet. In addition, most research on clustering of these health behaviors comes from high income countries (e.g. USA, UK, Australia, and Canada). As the cluster patterns may be unique to particular cultures [8, 12], there is a need to investigate this in low income (e.g. Kenya), lower-middle income (e.g. India), and upper-middle income countries (e.g. Brazil, China, and Colombia). There is also a lack of evidence regarding whether clusters of these health-related behaviors are only present in some sociodemographic groups [5, 8, 9, 16], or whether clusters of unhealthy behaviors appear to be more likely to occur among some sociodemographic groups and not others [11]. Previous findings showed that cluster differed by sex [6, 8, 9, 11, 17, 18], age [8, 11, 18], social class [11] and maternal education levels [8, 17]. The identification of which adolescents are more likely to engage in unhealthy behavior patterns allows us to understand which subgroups may be most at risk in terms of short- and long-term health outcomes [9]. The discrimination of population when associated to sociodemographic indicators, may help to better recognize and appropriately select strategies for obesity prevention [5]. The purpose of this study was to identify clustering of diet, PA and SB and association with sociodemographic variables among a national population-based sample of Brazilian adolescents. Such information might guide the development of health promotion programs, allowing a more targeted approach. Page 2 of 9 Meth (...truncated)