Background In recent years, researchers have begun applying a trajectory approach to identify homogeneous subgroups of physical activity (PA) in heterogeneous populations. This study systematically reviewed the articles identifying longitudinal PA trajectory classes and the related factors (e.g., determinants, predictors, and outcomes) in the general population during different life phases. Methods The included studies used finite mixture models for identifying trajectories of PA, exercise, or sport participation. Three electronic databases, PubMed (Medline), Web of Science, and CINAHL, were searched from the year 2000 to 13 February 2018. The study was conducted according to the PRISMA recommendations. Results Twenty-seven articles were included and organized into three age group: youngest (eleven articles), middle (eight articles), and oldest (eight articles). The youngest group consisted mainly of youth, the middle group of adults and the oldest group of late middle-aged and older adults. Most commonly, three or four trajectory classes were reported. Several trajectories describing a decline in PA were reported, especially in the youngest group, whereas trajectories of consistently increasing PA were observed in the middle and oldest group. While the proportion of persistently physically inactive individuals increased with age, the proportion was relatively high at all ages. Generally, male gender, being Caucasian, non-smoking, having low television viewing time, higher socioeconomic status, no chronic illnesses, and family support for PA were associated either with persistent or increasing PA. Conclusions The reviewed articles identified various PA subgroups, indicating that finite mixture modeling can yield new information on the complexity of PA behavior compared to studying population mean PA level only. The studies also provided novel information how different factors relate to changes in PA during life course. The recognition of the PA subgroups and their determinants is important for the more precise targeting of PA promotion and PA interventions. Trial registration PROSPERO registration number: CRD42018088120 . Electronic supplementary material The online version of this article (10.1186/s12889-019-6513-y) contains supplementary material, which is available to authorized users.

The purpose of this study was to investigate reciprocal relationships among students’ motor competence (MC) (leaping, throwing, catching, jumping skills), perceived physical competence, health‐related fitness (HRF) (20 m shuttle run, push‐up, abdominal muscles endurance tests) and objectively measured moderate‐to‐vigorous physical activity (MVPA). Participants included 422 Grade 5 Finnish children (246 girls). Two separate structural equation models investigated paths (a) from MC through both perceived physical competence and HRF to MVPA, and (b) from MVPA through both perceived physical competence and HRF to MC. Model 1 demonstrated an indirect path from MC through HRF to MVPA and a direct path from MC to perceived physical competence for both boys and girls. Additionally, model 1 revealed a direct path from perceived physical competence to MVPA for the girls and from MC to MVPA for the boys. MC, perceived physical competence, and HRF explained 13% of variance in MVPA for the girls and 25% for the boys. Model 2 indicated indirect paths from MVPA through perceived physical competence to MC and from MVPA through HRF to MC for both boys and girls. Additionally, a direct path from MVPA to MC was found in the boys’ group. MVPA, perceived physical competence, and HRF explained 48% of variance in MC for the girls and 53% for the boys. Results of this study provide preliminary support for the reciprocal nature of relationships among MC development, perceived physical competence, HRF and MVPA.

No abstract

Introduction: Despite substantial interest in the development of health behaviors, there is limited research that has examined the longitudinal relationship between physical activity (PA) and smoking trajectories from youth to adulthood in a Finnish population. This study aimed to identify trajectories of smoking and PA for males and females, and study the relationship between these trajectories from youth to adulthood. Methods: Latent profile analysis (LPA) was used to identify trajectories of smoking and PA separately for males and females among 3355 Finnish adults (52.1% females). Participants’ smoking and PA were assessed five to eight times over a 31-year period (3–18 years old at the baseline, 34–49 years at last follow-up). Multinomial logistic regression analysis was used to study the relationship between the trajectories of smoking and PA. Results: Five smoking trajectories and four to five PA trajectories were identified for males and females. Of the PA trajectory groups, the persistently active group were least likely to follow the trajectories of regular smoking and the inactive and low active groups were least likely to follow non-smoking trajectory group. Likewise, inactive (women only) and low active groups were less likely to belong to the non-smokers group. Conclusions: The study suggests that those who are persistently active or increasingly active have substantially reduced probabilities of being in the highest-risk smoking categories.

No abstract