Search

SUMMARY Background To explore at what age youth start using tobacco and/or nicotine products, which product is used first, product initiation sequences, and whether some socio-demographic characteristics are associated with substance use initiation. Methods Data were collected from an online questionnaire disseminated through social media and professional partners, targeting youth aged 14–25 in French-speaking Switzerland in August 2022. The final sample included 1362 participants. Respondents were asked whether they had already used cigarettes, e-cigarettes, hookah, snus, or cannabis at least once in their life (yes/no) and the age of the first time was asked for each substance when the answer was positive. Additionally, participants were asked about their substance use in the past 30 days. Respondents were classified according to age at onset of each tobacco/nicotine and cannabis product, and comparisons were made based on sociodemographic variables, including gender (cisgender female/cisgender male/transgender), perceived socioeconomic status (below average/average/above average), and age. Results Overall, in addition to those who reported no consumption, four distinct initiation profiles emerged from the classification. Cigarettes remain the most commonly used first product with an average age of 15.7 years. While cigarettes and hookah are present in all profiles, the order of first consumption varies from one profile to another, with cigarettes coming first in two profiles, hookah in one and e-cigarettes in the last. Furthermore, while the most common profile contains experimental consumption of the five products considered, some profiles do not contain cannabis, e-cigarettes and/or snus, for example. When divided by age groups, both 14-17-year-olds and 18-21-year-olds reported cigarette as their first product of initiation. Across the separate age groups (14–17, 18–21, 22–25), cigarettes consistently emerge as the primary used on average. Conclusions Cigarettes remain the first product to be used, but the younger the respondents, the earlier they start using e-cigarettes, and very close to cigarettes. Regarding current consumption patterns, e-cigarettes are becoming increasingly prominent.

Abstract Introduction Increasingly, literature has focused on community participation in places visited by persons living with and without dementia. Earlier research indicates that multiple factors, including socio-demographic aspects and risk perception may influence community participation. Aim and methods This cross-sectional, explorative study aims to inquire into how places visited, socio-demographic aspects and risks perception are associated with self-rated community participation for persons living with and without dementia (n = 70) in Switzerland. Data was collected through face-to-face interviews with questionnaires (ACT-OUT, MoCA, sociodemographic). First, we investigated whether the number of places visited was correlated with self-rated participation; then we added socio-demographic and risks perception factors with a bivariate analysis; and searched for a model using multinomial logistic regressions. Results For the group of participants living with dementia, risks of falling (p = .014) and of getting lost (p = .037) were significantly associated with self-rated participation. For the group of participants living without dementia, visiting places outside the home was significantly associated with self-rated participation, especially visiting places in domain D/places for recreational and physical activities (p = .005). Discussion and conclusions The results of exploring multiple factors and searching for a model highlights the complexity of community participation as a construct. Risks and visiting places for recreational and physical activities seem to play a role in self-rated participation. Mobile interviews might be better suited to gain in-depth understanding on community participation for persons living with dementia.

An accurate estimation of critical speed (CS) is important to accurately define the boundary between heavy and severe intensity domains when prescribing exercise. Hence, our aim was to compare CS estimates obtained by statistically appropriate fitting procedures, i.e., regression analyses that correctly consider the dependent variables of the underlying models. A second aim was to determine the correlations between estimated CS and aerobic fitness parameters, i.e., ventilatory threshold, respiratory compensation point, and maximal rate of oxygen uptake. Sixteen male runners performed a maximal incremental aerobic test and four exhaustive runs at 90, 100, 110, and 120% of the peak speed of the incremental test on a treadmill. Then, two mathematically equivalent formulations (time as function of running speed and distance as function of running speed) of three different mathematical models (two-parameter, three-parameter, and three-parameter exponential) were employed to estimate CS, the distance that can be run above CS (d′), and if applicable, the maximal instantaneous running speed (smax). A significant effect of the mathematical model was observed when estimating CS, d′, and smax (P < 0.001), but there was no effect of the fitting procedure (P > 0.77). The three-parameter model had the best fit quality (smallest Akaike information criterion) of the CS estimates but the highest 90% confidence intervals and combined standard error of estimates (%SEE). The 90% CI and %SEE were similar when comparing the two fitting procedures for a given model. High and very high correlations were obtained between CS and aerobic fitness parameters for the three different models (r ≥ 0.77) as well as reasonably small SEE (SEE ≤ 6.8%). However, our results showed no further support for selecting the best mathematical model to estimate critical speed. Nonetheless, we suggest coaches choosing a mathematical model beforehand to define intensity domains and maintaining it over the running seasons.

Adolescence is a period of life during which many people experiment with different kinds of legal and illegal substances. However, young people practising physical activities at a high level should avoid taking such substances. On the basis of a large sample of Swiss male recruits (C-SURF baseline data), we explore the consumption of substances among three subgroups of young adults, defined according to level of physical activity: high, medium, and low. Our results show that respondents classified into the high level of physical activity group went through the same experimentation processes with substances as respondents in the other groups, but that they reduced their overall consumption level, as indicated by measures regarding the last 12 months only. However, substantial differences are observed when we look at each substance separately. In particular, smokeless tobacco products are consumed more in the high group, and alcohol consumption is high in all groups. Physical activity, even at a high level, is not a protective factor against substance consumption. Therefore, physicians should not forget to investigate substance use among people with high levels of physical activity, especially since their consumption can (1) differ from the general population; and (2) have important consequences on their physical performance.

Missing data occur in almost all surveys; in order to handle them correctly it is essential to know their type. Missing data are generally divided into three types (or generating mechanisms): missing completely at random, missing at random, and missing not at random. The first step to understand the type of missing data generally consists in testing whether the missing data are missing completely at random or not. Several tests have been developed for that purpose, but they have difficulties when dealing with non-continuous variables and data with a low quantity of missing data. Our approach checks whether the missing data are missing completely at random or missing at random using a regression model and a distribution test, and it can be applied to continuous and categorical data. The simulation results show that our regression-based approach tends to be more sensitive to the quantity and the type of missing data than the commonly used methods.

High-order Markov chains are very useful for the analysis of complex temporal relationships, but they generally require a very high number of parameters [...]

When working with Markov chains, especially if they are of order greater than one, it is often necessary to evaluate the respective contribution of each lag of the variable under study on the present. This is particularly true when using the Mixture Transition Distribution model to approximate the true fully parameterized Markov chain. Even if it is possible to evaluate each transition matrix using a standard association measure, these measures do not allow taking into account all the available information. Therefore, in this paper, we introduce a new class of so-called "predictive power" measures for transition matrices. These measures address the shortcomings of traditional association measures, so as to allow better estimation of high-order models.

This article illustrates instrumental variable (IV) estimation by examining an unexpected finding of the research on career adaptability and job satisfaction. Theoretical and empirical arguments suggest that in the general population, people’s abilities to adapt their careers are beneficial to their job satisfaction. However, a recent meta-analysis unexpectedly found no effect when personality traits are controlled for. We argue that a reverse effect of job satisfaction on career adaptability, originating from affective tendencies tied to personality, might explain this null effect. Our argument implies that the estimates obtained with traditional ordinary least squares (OLS) regressions are biased by endogeneity, a correlation between an explanatory variable and the error term in a regression model. When experimental manipulations are impossible, IV estimations, such as two-stage least squares (2SLS) regressions, are one possible solution to the endogeneity problem. Analyzing three waves of data from a sample of 836 adults, the concurrent and time-lagged effect of job satisfaction on career adaptability was revealed to be more consistent than the reverse. Our results provide an explanation, rooted in affective dispositions, as to why recent meta-analytical estimates unexpectedly found that career adaptability does not predict job satisfaction at the interindividual level. We also discuss IV estimation in terms of its limits, weight the interpretation of its estimates against the temporality criterion for causal inference, and consider its possible extension to analyses of change.

Optimization of mixture models such as the mixture transition distribution (MTD) model is notoriously difficult because of the high complexity of their solution space. The best approach comprises combining features of two types of algorithms: an algorithm that can explore as completely as possible the whole solution space (e.g., an evolutionary algorithm), and another that can quickly identify an optimum starting from a set of initial conditions (for instance, an EM algorithm). The march package for the R environment is a library dedicated to the computation of Markovian models for categorical variables. It includes different algorithms that can manage the complexity of the MTD model, including an ad hoc hill-climbing procedure. In this article, we first discuss the problems related to the optimization of the MTD model, and then we show how march can be used to solve these problems; further, we provide different syntaxes for the computation of other models, including homogeneous Markov chains, hidden Markov models, and double chain Markov models.

The Mixture Transition Distribution (MTD) model used for the approximation of high-order Markov chains does not allow a simple calculation of confidence intervals, and computationnally intensive methods based on bootstrap are generally used. We show here how standard methods can be extended to the MTD model as well as other models such as the Hidden Markov Model. Starting from existing methods used for multinomial distributions, we describe how the quantities required for their application can be obtained directly from the data or from one run of the E-step of an EM algorithm. Simulation results indicate that when the MTD model is estimated reliably, the resulting confidence intervals are comparable to those obtained from more demanding methods.

QUESTION UNDER STUDY Switzerland has been receiving migrants of various origins for more than 50 years. The adoption of risk-taking behaviours among migrant youths is unclear. Moreover, when studied, migrant youths are rarely analysed according to whether they are first or second generation, or just young people with mixed origins. The aim of this study was to assess whether there are any differences between first-and second-generation immigrants, youths of mixed origins and their native peers in Switzerland concerning their engagement in risk behaviours. METHODS A total of 5834 youths from eleven post-mandatory schools in the canton of Fribourg (Switzerland) participated in the baseline survey of the GenerationFRee study, a longitudinal study to assess their lifestyle. Participants were divided by gender and by origin into: (a) natives: Swiss-born youths with Swiss-born parents, (b) first-generation migrants: foreign-born youths with foreign-born parents, (c) second-generation migrants: Swiss-born youths with foreign-born parents, (d) mixed-origin youths: Swiss-born youths with one Swiss-born parent and one foreign-born parent. Participants reported personal, family and school information, and attitudes towards eight risk behaviours. All significant variables at the bivariate level were included in a binary logistic regression. RESULTS The logistic regression showed that, compared with natives, first-and second-generation migrant boys were less likely to misuse alcohol. Boys of mixed origins were similar to migrants, although at the bivariate level they were more exposed to risk behaviours than were migrants. First-and second-generation migrant girls were less likely to misuse alcohol but three times more likely to be excessive Internet users. Girls of mixed origin were more likely to have their parents not living together and reported antisocial behaviours almost twice more often. CONCLUSIONS Our findings expose a lower engagement in risk behaviours among migrants. The migrant status in these two groups is clearly buffered if other control variables are considered. Thus, we can affirm that in the present study, migrants are not a high-risk population or not more at risk than the native group. Mixed origin youths showed higher risk behaviours than natives and migrants. Special attention should be given to this specific group, as they may be more vulnerable during adolescence.

Bullying is an issue of public health importance among adolescents worldwide. The present study aimed at explaining differences in bullying rates among adolescents in 15 low- and middle-income countries using globally comparable indicators of social and economic well-being. Using data derived from the Global School-based Health Survey, we performed bivariate analyses to examine differences in bullying rates by country and by bullying type. We then constructed a multi-level model using four fixed variables (age, gender, hunger and truancy) at the individual level, random effects at the classroom and school levels and four fixed variables at the country level (Gini coefficient, per capita Gross Domestic Project, homicide rate and pupil to teacher ratio). Bullying rates differed significantly by classroom, school and by country, with Egypt (34.2%) and Macedonia (3.6%) having the highest and lowest rates, respectively. Eleven-year-olds were the most likely of the studied age groups to report being bullied, as was being a male. Hunger and truancy were found to significantly predict higher rates of bullying. None of the explanatory variables at the country level remained in the final model. While self-reported bullying varied significantly between countries, the variance between classrooms better explained these differences. Our findings suggest that classroom settings should be considered when designing approaches aimed at bullying prevention.

Test–retest is a concept that is routinely evaluated during the validation phase of many measurement tools. However, this term covers at least two related but very different concepts: reliability and agreement. Reliability is the ability of a measure applied twice upon the same respondents to produce the same ranking on both occasions. Agreement requires the measurement tool to produce twice the same exact values. An analysis of research papers showed that the distinction between both concepts remains anything but clear, and that the current practice is to evaluate reliability only, generally on the basis of the sole Pearson’s correlation. This practice is very problematic in the context of longitudinal studies because it becomes difficult to determine whether a difference between two successive observations is attributable to a real change of the respondents or only to the characteristics of the measurement tool, which then leads to a possible misinterpretation of the results. More focus should be given on the real interpretation of linear correlation, and when agreement is required in addition to reliability, then correct alternative, such as the Bland–Altman plot, should be more generally used.

OBJECTIVE: To assess whether problematic internet use is associated with somatic complaints and whether this association remains when checking for internet activity among a random sample of adolescents living in the canton of Vaud, Switzerland. METHODS: Cross-sectional survey of 3,067 8th graders (50.3% females) divided into average (n = 2,708) and problematic (n = 359) Internet users and compared for somatic complaints (backache, overweight, headaches, musculoskeletal pain, sleep problems and sight problems) controlling for sociodemographic and internet-related variables. Logistic regressions were performed for each complaint and for all of them simultaneously controlling variables significant at the bivariate level. RESULTS: At the multivariate level, when taken separately, problematic internet users were more likely to have a chronic condition (adjusted odds ratio [aOR] with 95% CI: 1.58 [1.11:2.23]) and to report back pain (aOR: 1.46 [1.04:2.05]), overweight (aOR: 1.74 [1.03:2.93]), musculoskeletal pain (aOR: 1.36 [1.00:1.84]) and sleep problems (aOR: 2.16 [1.62:2.88]). When considered in the full model, only sleep problems remained significant (aOR: 2.03 [1.50:2.74]). CONCLUSIONS: Our results confirm that problematic internet users report health problems more frequently, with lack of sleep being the most strongly associated and seeming to act as mediator regarding the other ones. Clinicians should remember to screen for excessive internet use their patients complaining of sleep-related problems, back or musculoskeletal pain or overweight. Clinicians should advise parents to limit the amount of time their adolescent children can spend online for leisure activities. Furthermore, limiting the number of devices used to connect to the internet could help warrant enough sleeping time.

Sports-practicing youths are at an elevated risk for alcohol use and misuse. Although much attention has recently been given to depicting subgroups facing the greatest threats, little evidence exists on the contexts in which their drinking takes place. Using data from a cross-sectional study on youth sports participation and substance use in the French-speaking part of Switzerland, this study focused on the social contexts associated with hazardous drinking of 894 sports-practicing adolescents aged 16 to 20. Divided between those who had been drunk in the last month (hazardous drinkers, n = 315) and those who had not (n = 579), sports-practicing adolescents were compared on reported gatherings (sports-related, sports-unrelated, mixed) likely linked to their drinking behaviour. Mixed social contexts, followed by sports-unrelated ones, were reported as the most common context by both male and female youths who practiced sports. After controlling for several possible confounders, male hazardous drinkers were more than 3 times more likely to report sports-unrelated social contexts as the most common, compared to sport-related ones, while females were more than 7 times more likely to do so. Our findings seem to indicate that, rather than focusing only on sports-related factors, prevention of alcohol misuse among sports-practicing youths should also pay attention to the social contextualisation of their hazardous drinking.

MARCH is a free software for the computation of different types of Markovian models including homogeneous Markov Chains, Hidden Markov Models (HMMs) and Double Chain Markov Models (DCMMs). The main characteristic of this software is the implementation of a powerful optimization method for HMMs and DCMMs combining a genetic algorithm with the standard Baum-Welch procedure. MARCH is distributed as a set of Matlab functions running under Matlab 5 or higher on any computing platform. A PC Windows version running independently from Matlab is also available.