Clustering of multiple health-risk factors among vocational education students: a latent class analysis

Abstract

Physical and mental health risks often commence during young adulthood. Vocational education institutions are an ideal setting for understanding how health-risks cluster together in students to develop holistic multiple health-risk interventions. This is the first study to examine clustering of tobacco smoking, fruit intake, vegetable intake, alcohol consumption, physical inactivity, overweight/obesity, depression, and anxiety in vocational education students and the socio-demographic characteristics associated with cluster membership. A cross-sectional survey with vocational education students (n = 1134, mean age = 24.3 years) in New South Wales, Australia. Latent class analysis identified clusters and latent class regression examined characteristics associated with clusters. Four clusters were identified. All clusters had moderate inadequate fruit intake and moderate overweight/obesity. Cluster 1 (13% of sample) had “high anxiety, high inadequate vegetable intake, low tobacco, and low alcohol use.” Cluster 2 (16% of sample) had “high tobacco smoking, high alcohol use, high anxiety, high depression, and high inadequate vegetable intake.” Cluster 3 (52% of sample) had “high risky alcohol use, high inadequate vegetable intake, low depression, low anxiety, low tobacco smoking, and low physical inactivity.” Cluster 4 (19% of sample) was a “lower risk cluster with high inadequate vegetable intake.” Compared to cluster 4, 16–25-year-olds and those experiencing financial stress were more likely to belong to clusters 1, 2, and 3. Interventions for vocational education students should address fruit and vegetable intake and overweight/obesity and recognize that tobacco use and risky alcohol use sometimes occurs in the context of mental health issues.

Clustering of chronic disease risks among people accessing community mental health services

“…Additionally, females and older people were more likely to be in Cluster 3 than Cluster 1, compared to males and those aged 18–34, respectively. As Cluster 3 exhibited the relatively lowest probability for harmful acute alcohol consumption and relatively highest probabilities for both types of inadequate activity, these findings are consistent with previous research demonstrating females and older people are less likely to binge drink ( Pettigrew et al, 2021 , Atorkey et al, 2021a , Atorkey et al, 2021b , Shaw and Agahi, 2012 , Lee et al, 2012 , Griffin et al, 2014 , Paul et al, 2016 , Noel et al, 2013 ) as well as engage in adequate physical activity ( Alley et al, 2017 , Bennie et al, 2016 ) than males and younger people, respectively. Understanding characteristics associated with certain clusters may provide additional considerations for tailoring multi-risk interventions.…”

“…Clustering of such risks is supported by previous research suggesting that physical activity impacts diet and weight ( Beaulieu et al, 2016 , Fenton et al, 2021 , Höchsmann et al, 2020 ), a potentially more salient association for people with a mental health condition due to health promoting barriers such as medication induced cravings and low motivation ( Bailey et al, 2018 , Scott and Happell, 2011 ). The probability of tobacco smoking in Cluster 3 (0.46) is typically higher than previous studies which have identified ‘unhealthy’ clusters that are not characterised by smoking (probabilities ranging from 0.04 to 0.29) ( Atorkey et al, 2021b , Hutchesson et al, 2021 , Oftedal et al, 2019 , Pettigrew et al, 2021 ), evidence of tobacco smoking being a particularly prevalent risk among this population group, likely due to the perceived role of tobacco smoking in mental illness symptom management ( Bailey et al, 2018 , Keller-Hamilton et al, 2019 ).…”

A latent class analysis of health risk behaviours in the UK Police Service and their associations with mental health and job strain

“…This is in line with the few existing studies which demonstrate that individuals with poor mental health are more likely to be assigned to clusters of high health risk behaviours [ 12 , 14 , 60 ]. One study of vocational education students explored the clustering of physical and mental health risks, finding a class of high anxiety, high depression, and multiple health risk behaviours (smoking, risky alcohol use, poor diet, physical inactivity) [ 27 ]. A study independently examining associations between health risk behaviours and mental health, found that poor mental health was linked with smoking, low fruit and vegetable intake, and abstinence [ 61 ].…”

“…However, it is not known how health risk behaviours cluster together in UK police employees. Latent Class Analysis (LCA) is a statistical technique used to identify distinct classes based on responses to multiple variables, and has previously been used to determine classes of health risk behaviours in adolescents [ 26 ], vocational education students [ 27 ], older adults [ 28 ] and the UK general population [ 29 ]. This study aims to utilise LCA to identify classes of health risk behaviours (alcohol use, fruit and vegetable intake, red meat consumption, smoking, physical activity) in UK police employees, and to determine their associations with mental health (depression, anxiety, PTSD) and job strain (high, low, active, passive).…”

Polysubstance Use and Correlations with Psychosocial and Health Risk Behaviours a Latent Profile Analysis Among More than 95,000 Norwegian Adolescents During the COVID-19 Pandemic

“…Other studies show that low levels of parental monitoring (14), high incidences of mental health problems (9), high levels of conduct problems (15), and physical pain (i.e., headache, back pain) were associated with polysubstance use (1). Finally, although risky health behaviours, such as a poor diet and low levels of physical activity, have rarely been included in LPA polysubstance use studies, a few studies found that high levels of alcohol and tobacco use are related to a low consumption of fruit and vegetables (16,17) and low physical activity levels a sedentary lifestyle (17). Although these studies found an overall relationship between these variables and polysubstance use, few studies have included a wide array of predictors using the same sample.…”