Predicting anxiety, depression, and insomnia among Bangladeshi university students using tree-based machine learning models

Background and Aims

Mental health problem is a rising public health concern. People of all ages, specially Bangladeshi university students, are more affected by this burden. Thus, the objective of the study was to use tree-based machine learning (ML) models to identify major risk factors and predict anxiety, depression, and insomnia in university students.

Methods

A social media-based cross-sectional survey was employed for data collection. We used Generalized Anxiety Disorder (GAD-7), Patient Health Questionnaire (PHQ-9) and Insomnia Severity Index (ISI-7) scale for measuring students' anxiety, depression and insomnia problems. The tree-based supervised decision tree (DT), random forest (RF) and robust eXtreme Gradient Boosting (XGBoost) ML algorithms were used to build the prediction models and their predictive performance was evaluated using confusion matrix and receiver operating characteristic (ROC) curves.

Results

Of the 1250 students surveyed, 64.7% were male and 35.3% were female. The students' ages ranged from 18 to 26 years old, with an average age of 22.24 years (SD = 1.30). Majority of the students (72.6%) were from rural areas and social media addicted (56.6%). Almost 83.3% of the students had moderate to severe anxiety, 84.7% had moderate to severe depression and 76.5% had moderate to severe insomnia problems. Students' social media addiction, age, academic performance, smoking status, monthly family income and morningness-eveningness are the main risk factors of anxiety, depression and insomnia. The highest predictive performance was observed from the XGBoost model for anxiety, depression and insomnia.

Conclusion

The study findings offer valuable insights for stakeholders, families and policymakers enabling a more profound comprehension of the pressing mental health disorders. This understanding can guide the formulation of improved policy strategies, initiatives for mental health promotion, and the development of effective counseling services within university campus. Additionally, our proposed model might play a critical role in diagnosing and predicting mental health problems among Bangladeshi university students and similar settings.

The Impact of Mental Health Predictors of Internet Addiction among Pre-Service Teachers in Ghana

This study examined the prevalence of addictive Internet behavior and its links with mental health among pre-service teachers in Ghana. A descriptive, correlational design was employed with 405 pre-service teachers from colleges of education and a public university in Ghana participating in this study. The sample completed a sociodemographic survey about loneliness, life satisfaction, depression, self-esteem, and the Internet addiction scales (abridged form). The results revealed that there was a significant relationship between pre-service teachers' Internet addiction, depression, life satisfaction, and loneliness; however, depression was the least influential factor in addictive Internet use. Additionally, there was a statistically significant nexus between self-esteem, loneliness, depression, and life satisfaction. In addition, all the above-mentioned variables were discovered to explain 56.3% of the absolute variance in addiction to the Internet. Among the variables linked with Internet addiction and its dimensions, loneliness appeared to be the most significant. Institutional coping programs with Internet addiction should be established within the scope of the university administration, supporting pre-service teachers' mental health. Finally, the development of awareness campaigns on the menaces associated with Internet usage and mental health through extracurricular programs is recommended.

Prevalence of Smartphone Addiction and Its Association with Sociodemographic, Physical and Mental Well-Being: A Cross-Sectional Study among the Young Adults of Bangladesh

Smartphones have made our lives easier and have become indispensable for everyday life; however, their uncontrolled and excessive use can trigger "smartphone addiction" (SA). SA is a rising public health issue, particularly among young people around the world. There is a dearth of empirical research about SA and its impacts on young adults, particularly in developing countries such as Bangladesh. This cross-sectional study is an attempt to fill this gap. The data were collected from 440 eligible young adults in Bangladesh using an online survey between July 2021 and February 2022. Study results revealed that 61.4% of the young adults were addicted to their smartphone. Logistic regression analysis showed that being male, aged ≤25, unemployed and living with a large family size (≥8) were the significant sociodemographic predictors of SA. Smartphone-addicted participants were more likely to be less physically active, suffer from insomnia, be overweight or obese and use their phones while driving, walking and eating. In addition, the SA group were more likely to have physical and mental well-being problems. This study brings to light significant implications for policy makers and indicates a need for an SA community awareness programme which aims to reduce SA at the societal level.

Impact of a Digital Intervention for Literacy in Depression among Portuguese University Students: A Randomized Controlled Trial

Digital interventions are important tools to promote mental health literacy among university students. “Depression in Portuguese University Students” (Depressão em Estudantes Universitários Portugueses, DEEP) is an audiovisual intervention describing how symptoms can be identified and what possible treatments can be applied. The aim of this study was to evaluate the impact of this intervention. A random sample of 98 students, aged 20–38 years old, participated in a 12-week study. Participants were recruited through social media by the academic services and institutional emails of two Portuguese universities. Participants were contacted and distributed into four study groups (G1, G2, G3 and G4): G1 received the DEEP intervention in audiovisual format; G2 was given the DEEP in text format; G3 received four news articles on depression; G4 was the control group. A questionnaire was shared to collect socio-demographic and depression knowledge data as a pre-intervention method; content was then distributed to each group following a set schedule; the depression knowledge questionnaire was then administered to compare pre-intervention, post-intervention and follow-up literacy levels. Using the Scheffé and Least Significant Difference (LSD) multiple comparisons test, it was found that G1, which received the DEEP audiovisual intervention, differed significantly from the other groups, with higher depression knowledge scores in post-intervention stages. The DEEP audiovisual intervention, compared to the other formats used (narrative text format; news format), proved to be an effective tool for increasing depression knowledge in university students.

How Has the COVID-19 Pandemic Impacted Internet Use Behaviors and Facilitated Problematic Internet Use? A Bangladeshi Study

Background

The COVID-19 pandemic-related "stay-at-home" and confinement orders has led individuals to be more engaged with technology use (eg, internet use). For a minority of individuals, excessive use can become problematic and addictive. However, the investigation of problematic internet use in the COVID-19 context is only just emerging. Therefore, the present study investigated the changes in internet use behaviors and addiction rates in comparison with prior Bangladeshi studies.

Methods

An online cross-sectional study was carried out among a total of 601 Bangladeshi students between October 7 and November 2, 2020. The survey included questions relating to socio-demographic, behavioral health, online use behaviors, and psychopathological variables.

Results

A quarter of the participants (26%) reported having low levels of internet addiction, whereas 58.6% were classed as having moderate internet addiction and 13% severe internet addiction. A total of 4% of the sample were classed as being at risk of severe internet dependency (ie, scoring over ≥80 on IAT). Risk factors for internet addiction included smartphone addiction, Facebook addiction, depression, and anxiety. However, the final hierarchical regression model comprising all variables explained a total of 70.6% variance of problematic internet use.

Conclusion

Based on the present findings, it is concluded that individuals are at elevated risk of problematic internet use like other psychological impacts that have been reported during the COVID-19 pandemic. Therefore, risk-reducing measures and healthy control use strategies should be implemented for vulnerable individuals.

Smartphone Addiction and Quality of Sleep among Indian Medical Students

Background: Smartphone is an indispensable miracle of artificial intelligence in the hands of global netizens. Medical students overburdened and sleep-deprived due to the curricular demands can have serious health effects due to further sleep deprivation caused by problematic smartphone use, affecting their work proficiency as practicing physicians. This study aimed to evaluate the association of smartphone addiction and quality of sleep among medical students across all semesters, along with other background variables.Methodology: This cross-sectional study was conducted on 224 medical students of a tertiary care teaching hospital in North India using a self-administered questionnaire with four parts -Socio-demographic characteristics, General health questionnaire (GHQ-12), Smartphone addiction scale-short version (SAS-SV), and Pittsburgh sleep quality index (PSQI).Results: The prevalence of smartphone addiction is was found out to be 33.33% in females and 46.15% in males. In the study, 63.39% were poor sleepers as assessed by their PQSI scores, and 62.05% reported poor health status as per their GHQ scores. There was a positive correlation between overall PQSI scores and smartphone use duration per day, SAS-SV scores, and GHQ scores.Conclusion: The high prevalence of excessive smartphone usage among medical students is a cause for concern and is detrimental to their health and sleep quality. The research addresses current lacunae in correlating smartphone addiction with smartphone usage before sleeping and right after waking up. Risky behavior adoption and pervasive mood changes associated with excessive smartphone use are addressed with equal representation across semesters. The study recommends increasing self-awareness for help-seeking to regulate smartphone usage and providing counseling services for students in their formative years.

Problematic internet use among young and adult population in Bangladesh: Correlates with lifestyle and online activities during the COVID-19 pandemic

BACKGROUND

Although internet use can boost individuals' quality of life in various aspects, activities on the internet (e.g., gambling, video gaming, social media use, pornography use, etc.) can be used as coping strategy to deal with psychological stressors and mood states (e.g., fear, anxiety, depression) particularly during the global COVID-19 pandemic.

OBJECTIVES

The present study assessed problematic internet use (PIU) among Bangladeshi youth and adults in Bangladesh and examined its correlation with lifestyle and online activities during the COVID-19 pandemic.

METHODS

An online cross-sectional survey was utilized between May and June 2020 comprising 13,525 Bangladeshi individuals (61.3% male; age range 18-50 years; mean age 23.7 years) recruited from various online platforms. The self-report survey included questions concerning socio-demographics, lifestyle, and online activities during the COVID-19 pandemic, as well as psychometric scales such as the nine-item Internet Disorder Scale-Short Form (IDS9-SF).

RESULTS

Utilizing hierarchical regression analysis, problematic internet use was significantly and positively associated with those who were younger, having a higher level of education, living with a nuclear family, engaging in less physical exercise, avoiding household chores, playing online videogames, social media use, and engaging in recreational online activities.

CONCLUSIONS

Excessive internet use appears to have been commonplace during the COVID-19 pandemic period and young adults were most vulnerable to problematic internet use.

Understanding the Phenomenon of Binge-Watching-A Systematic Review

Binge-watching is a relatively new behaviour pattern whose popularity has been increasing since 2013, ultimately to become one of the most popular ways of spending free time, especially among young people. However, there is still a dearth of research on this phenomenon. The aim of this study is to present the current understanding and psychological conditions of binge-watching, as provided in the research papers published between 2013 and 2020. This systematic review, including 28 articles, addresses different approaches to defining this behaviour, diverse motivations, personality traits, and risks of excessive binge-watching. Its results imply that there are two perspectives in understanding binge-watching. The first is related to entertainment, positive emotions, cognition, and spending free time. However, the second perspective emphasises the negative outcomes of excessive binge-watching and symptoms of behavioural addiction. There is undoubtedly a need for further research to be conducted on diversified populations to reach more profound understanding of binge-watching behaviour patterns.

Does loneliness, self-esteem and psychological distress correlate with problematic internet use? A Bangladeshi survey study

BACKGROUND

Researchers have claimed that problematic internet use (PIU) and internet addiction (IA) is a global mental health problem. However, little research has addressed this issue in Bangladesh and no previous study has examined the relationship between PIU and potential psychological risk factors.

AIM

The present study examined the prevalence of PIU and its associated risk factors including socio-demographic variables, internet use behaviors, and other psychological variables including loneliness, self-esteem, and psychological distress.

METHODS

A cross-sectional study was conducted among Bangladeshi students (N = 605). Measures included the Internet Addiction Test (IAT), Rosenberg's Self-Esteem Scale, the UCLA Loneliness Scale, and the General Health Questionnaire.

RESULTS

Using a cut-off score of ≤60 (out of 100) on the IAT, the prevalence of PIU was 16.5% (n = 100). Regression analysis showed that sleep disturbance (more or less than 6-7 hours sleeping time), not using internet for academic purposes, and online chatting were risk factors for PIU. Loneliness and psychological distress were positively correlated with PIU, whereas self-esteem was negatively correlated.

CONCLUSIONS

The present study highlights that PIU among Bangladeshi university students is an issue of concern and that targeted prevention is needed among emerging adults to help overcome the potentially negative effects of problematic internet use.

Adolescent problematic internet use and parental mediation: A Bangladeshi structured interview study

Internet-related problems such as excessive internet use, problematic internet use (PIU), and internet addiction, are becoming increasingly studied among Bangladeshi adult students, but there has been little research among adolescents. In Bangladesh, there has been no research examining the role of parental mediation in their children's internet use. Therefore, the present structured interview study investigated Bangladeshi adolescent PIU and its associated socio-demographics, internet use behaviors, and the parental mediation role among 350 high school students residing in Dhaka. The results showed that 84 of adolescents (24.0%) were classified as having PIU (cut-off score of ≥ 50 on the Internet Addiction Test) and nine adolescents (2.6%) were classified as having a severe dependency on the internet (cut-off score of >80 on the Internet Addiction Test). According to hierarchical regression analysis, significant PIU correlates included lower academic results, both parents' lower education, mother working outside the home, more than four days' weekly internet use, more than two hours daily internet use, and active mediation. Additionally, internet use behaviors (i.e., internet use locations, devices, purposes, and applications) and parental internet mediation dimensions other than active mediation (i.e., restrictive mediation, active mediation internet safety, monitoring, and technical mediation) were significantly related to PIU in t-tests and correlation analysis respectively. However, they were non-significant in the hierarchical regression analysis when included into equation altogether. The present study's findings will be helpful in developing country-level policymaking decisions and facilitating future research in the country.

Relationship between internet addiction and depression among Japanese university students

BACKGROUND

Internet addiction (IA) has various adverse effects. We sought to elucidate the relationship between IA and depression among university students and to identify factors associated with IA.

METHODS

Anonymous, self-administered questionnaires were distributed to 5,261 students and comprised basic characteristics, lifestyle habits, anxieties, the Internet Addiction Test (IAT), and the Center for Epidemiological Self-Depression Scale.

RESULTS

Responses were obtained from 4,490 students (response rate: 85.3%). After excluding those with missing responses, 3,251 participants were analyzed (valid response rate: 61.8%). Logistic regression analysis with severity of IA as the independent variable and depression as the dependent variable revealed that the odds ratio (OR) for depression increased with severity of IA (mild addiction: OR=2.87, 95% confidence interval [CI]=2.45-3.36; severe addiction: OR=7.31, 95% CI=4.61-11.61). In a logistic regression analysis with mobile phone use as the independent variable and IA as the dependent variable, the highest OR was for message board use (OR=3.74, 95% CI=2.53-5.53) and the lowest OR was for use of LINE instant messenger (OR=0.59, 95% CI=0.49-0.70). Logistic regression analysis with academic department as the independent variable and internet addiction as the dependent variable revealed high ORs for the humanities department (OR=1.59, 95% CI=1.18-2.16) and fine arts department (OR=1.55, 95% CI=1.07-2.23).

LIMITATIONS

The main limitations were the cross-sectional design, low valid response rate, single-university setting, and possible social desirability bias.

CONCLUSIONS

Our results suggest a relationship between IA and depression in university students. IA tendency differed according to mobile phone use and academic department, suggesting these factors are associated with IA.

Problematic internet use in Bangladeshi students: The role of socio-demographic factors, depression, anxiety, and stress

Problematic Internet Use (PIU) has become a concern for public mental health across the globe. However, there are few studies assessing PIU in Bangladesh. The present cross-sectional study estimated the prevalence rate of PIU and its associated risk factors among 405 university students in Bangladesh between June and July 2018. The measures included sociodemographic questions, internet and health-related variables, the Internet Addiction Test (IAT) and the Depression, Anxiety and Stress Scale (DASS-21). The prevalence of PIU was 32.6% among the respondents (cut-off score of ≥50 on the IAT). The prevalence of PIU was higher in males in comparison to females, although the difference was not statistically significant. Internet-related variables and psychiatric comorbidities were positively associated with PIU. From the unadjusted model, more frequent use of the internet and more time spent on the internet were identified as strong predictors of PIU, whereas the adjusted model showed depressive symptoms and stress only as strong predictors of PIU. It is hoped this preliminary study will facilitate further study on PIU along with other psychiatric disorders in Bangladesh.

Prevalence and Psychiatric Risk Factors of Excessive Internet Use among Northern Bangladeshi Job-Seeking Graduate Students: A Pilot Study

Although internet addiction (IA) has been the subject of research in many countries, it is poorly studied in Bangladesh, and previous studies in the country have not examined potential mental health risk factors in relation to the development of IA. The present pilot study attempted to assess the prevalence and associated risk factors of IA among a convenience sample of recently graduated university students actively looking for a job (N = 284) from July to October 2018 in Rajshahi (a city in Bangladesh). The measures included socio-demographic and behavioral variables, the Internet Addiction Test (IAT), and the Depression, Anxiety, and Stress Scale. Surprisingly, no cases of IA were identified in the sample (i.e., the prevalence rate of IA was 0%) which may have been a consequence of the specific population sampled. A total of 3.9% of the participants (n = 11) were classified as excessive users (≥ 60 in IAT). Using regression analysis, the risk factors for excessive internet use were spending more time online (> 5 h daily) and experiencing psychiatric conditions (i.e., depression, stress). It is hoped the present pilot study will facilitate further study in Bangladesh with a more in-depth focus on the socio-demographic and psychiatric morbidities in excessive and problematic internet use. It is also speculated that job-seeking may be a situational protective risk factor for developing IA.

Internet Addiction Through the Phase of Adolescence: A Questionnaire Study

BACKGROUND

Adolescents increasingly use the Internet for communication, education, entertainment, and other purposes in varying degrees. Given their vulnerable age, they may be prone to Internet addiction.

OBJECTIVE

Our aim was to identify possible differences in the purpose of Internet use among adolescents with respect to age subgroup, country of residence, and gender and the distribution of Internet addiction across age subgroups. Another aim was to determine if there is a correlation between the purpose of Internet use and age and if this interaction influences the level of addiction to the Internet.

METHODS

The study included a simple random sample of 1078 adolescents-534 boys and 525 girls-aged 11-18 years attending elementary and grammar schools in Croatia, Finland, and Poland. Adolescents were asked to complete an anonymous questionnaire and provide data on age, gender, country of residence, and purpose of Internet use (ie, school/work or entertainment). Collected data were analyzed with the chi-square test for correlations.

RESULTS

Adolescents mostly used the Internet for entertainment (905/1078, 84.00%). More female than male adolescents used it for school/work (105/525, 20.0% vs 64/534, 12.0%, respectively). Internet for the purpose of school/work was mostly used by Polish adolescents (71/296, 24.0%), followed by Croatian (78/486, 16.0%) and Finnish (24/296, 8.0%) adolescents. The level of Internet addiction was the highest among the 15-16-year-old age subgroup and was lowest in the 11-12-year-old age subgroup. There was a weak but positive correlation between Internet addiction and age subgroup (P=.004). Male adolescents mostly contributed to the correlation between the age subgroup and level of addiction to the Internet (P=.001).

CONCLUSIONS

Adolescents aged 15-16 years, especially male adolescents, are the most prone to the development of Internet addiction, whereas adolescents aged 11-12 years show the lowest level of Internet addiction.