Understanding Youth Health Risk Behaviours in Indonesia through a Structural Model of Social and Cognitive Determinants: A Longitudinal Analysis

Understanding Youth Health Risk Behaviours in Indonesia through a Structural Model of Social and Cognitive Determinants: A Longitudinal Analysis | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Understanding Youth Health Risk Behaviours in Indonesia through a Structural Model of Social and Cognitive Determinants: A Longitudinal Analysis Gisely Vionalita, Nathan Hughes, Alvaro Martinez-Perez This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7244173/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 9 You are reading this latest preprint version Abstract Background: Cognitive development during adolescence plays a critical role in shaping decision-making and self-regulation. However, its role in influencing health risk behaviours particularly as a pathway linking early-life social conditions to later health outcomes remains underexplored in middle-income settings. This study investigates how cognitive status both directly and indirectly influences health risk behaviours among Indonesian youth, with a focus on the mediating role of cognition between early social structure and later health behaviours. Methods: Using longitudinal data from the Indonesia Family Life Survey (2007–2014), which tracked individuals aged 15–30, we applied structural equation modelling (SEM) to test three hypotheses: (1) the direct effect of adolescent cognitive status on health risk behaviours in young adulthood; (2) its mediating role between early-life social structure and risk behaviour; and (3) its contribution to behavioural change over time in smoking and physical activity. Results: Higher cognitive status in adolescence was significantly associated with lower engagement in health risk behaviours such as smoking, unhealthy diet, and physical inactivity in young adulthood (standardized coefficient = -0.12, p < 0.001). Cognitive status partially mediated the relationship between early social advantage and health risk behaviours (indirect effect = -0.06, p < 0.001), while the direct effect of social structure remained significant (direct effect = -0.08, p < 0.01). Although cognitive status did not independently predict behavioural change over time, its indirect influence through social structure remained substantial (indirect effect = -0.22, p < 0.001).Keywords: health risk behaviour, adolescents, youth Indonesia, cognitive status, social structure. Conclusions: Cognitive development plays a dual role as a product of social conditions and a protective factor against health risk behaviours. In rapidly developing countries like Indonesia, policies that enhance cognitive development alongside efforts to improve social-economic conditions may be particularly effective in supporting healthier behaviours during the transition from adolescence to young adulthood. cognitive social structure adolescence health risk-behaviour structural modelling longitudinal study Figures Figure 1 Figure 2 Figure 3 Introduction Adolescence and young adulthood represent a critical developmental life stage during which individuals undergo significant cognitive maturation while also becoming increasingly exposed to health risk behaviours, including smoking, physical inactivity, and unhealthy dietary patterns. These years are marked by both vulnerability and opportunity: cognitive development including memory, abstract reasoning, and executive function enables adolescents to make complex decisions, anticipate long-term consequences, and regulate impulses (Junger & van Kampen, 2010 ; Whalley & Deary, 2001 ). At the same time, adolescents are influenced by shifting social environments, including peer pressure, changing family dynamics, and community expectations, which shape behavioural patterns that often persist into adulthood (Arnett, 2000 ; Steinberg, 2008 ). In high-income countries, a growing body of evidence has shown that adolescents with stronger cognitive function are less likely to engage in health risk behaviours such as smoking, substance use, and physical inactivity (Batty et al., 2007 ; Clark et al., 2007 ; Daly et al., 2022 ; Hackman et al., 2010 ). Cognitive abilities particularly those related to memory and reasoning have been linked to improved decision-making and better health-related judgement (Reyna & Farley, 2006 ). However, while this evidence base is well developed in high-income contexts, far less is known about how these relationships manifest in upper-middle-income countries, where social and economic transitions, regional disparities, and structural inequalities may differently shape both cognitive development and health behaviour trajectories (Patton et al., 2016 ; Viner et al., 2012 ). Although Indonesia is now classified as an upper-middle-income country, it continues to face uneven development, particularly in access to education and healthcare services. Youth in Indonesia, who comprise approximately 24% of the national population (aged 10–24) (UNFPA, 2025 ), navigate multiple layers of socio‑cultural and economic transitions that are likely to influence both their cognitive and behavioural outcomes. In this context, understanding the pathways through which cognitive development mediates the relationship between early social advantage and health behaviours is essential for designing effective public health interventions. The role of cognitive status in shaping adolescent health behaviour is hypothesised to be twofold. First, cognitive skills may exert a direct protective effect, enabling individuals to evaluate health risks more effectively and make informed behavioural choices (Costa et al., 1991 ; Reyna & Farley, 2006 ). Second, cognitive development may act as a mediating mechanism through which early-life structural conditions such as stable family conditions, higher parental education, and community support translate into healthier adolescent behaviour (Brieant et al., 2021 ; Hackman et al., 2010 ). Life-course theory suggests that the structural conditions experienced during adolescence leave lasting imprints on developmental outcomes, including cognitive capacity, which in turn influence future health trajectories (Halfon & Hochstein, 2002 ; Shonkoff et al., 2009 ). This framework is particularly relevant in the Indonesian context. Evidence shows that Indonesian adolescent males have among the highest smoking prevalence rates globally (Ng et al., 2014 ), while adolescent girls face gendered barriers in physical activity and food autonomy (Septiono et al., 2022 ). At the same time, disparities in family stability, education quality, and access to youth support services are likely to contribute to wide variability in both cognitive and behavioural outcomes across socioeconomic groups. Despite these dynamics, there is a notable lack of empirical research in upper-middle-income countries on how cognitive function interacts with structural factors to shape health behaviours during the transition from adolescence to adulthood. This study addresses that gap by analysing longitudinal Indonesian survey data to test a conceptual model in which cognitive status functions both as a direct predictor and a mediator of health risk behaviours. Drawing on life-course theory and the social determinants of health framework, this study uses structural equation modelling (SEM) to examine how adolescent cognitive function measured through verbal memory and fluid reasoning interacts with early social inequalities to shape health risk behaviours and behavioural changes by early adulthood. Specifically, the study addresses two research questions; (1) what is the association between cognitive status and health risk-taking behaviour among adolescents and young adults in Indonesia? And (2) does cognitive status mediate the relationship between early-life social structure and health risk-taking behaviour? The findings will seek to inform youth-focused public health strategies by providing evidence on how cognitive development might serve as a protective mechanism in reducing risk behaviours, particularly in rapidly developing contexts such as Indonesia. Methods Study Design and Data Source The aim of this study was to investigate the direct and indirect effects of cognitive status on health risk-taking behaviour among Indonesian adolescents and young adults, using longitudinal structural equation modelling. This study used data from the Indonesia Family Life Survey (IFLS) a longitudinal, nationally representative household survey spanning five waves from 1993 to 2014. For this analysis, we used data from Wave 4 (2007) and Wave 5 (2014) to examine the relationships between early-life social structure, cognitive function, and health risk-taking behaviours in adolescence and young adulthood. The IFLS employed a multistage stratified sampling method and included 13 provinces representing approximately 83% of the Indonesian population. Respondents aged 15–22 in 2007 were followed up as young adults (aged 23–30) in 2014. The survey included a wide array of health, economic, cognitive, and psychosocial measures collected via interviewer-administered and self-completed questionnaires (Strauss et al., 2016 ). Sample Selection We restricted the sample to individuals aged 15–22 years in 2007 and 23–30 years in 2014, yielding a total of 11,539 eligible participants. A balanced panel dataset was created by retaining respondents who participated in both waves, resulting in a final analytical sample of 2,525 individuals. While this resulted in attrition (~ 57%), unbalanced panel data were retained for supplementary and robustness analyses to enhance generalisability. Measurement of Key Constructs All variables were derived from the Indonesia Family Life Survey (IFLS) questionnaire and mapped to key theoretical constructs. Latent constructs were defined using validated survey items consistent with prior studies and grounded in the life-course framework (see Table 1 ). Table 1 Measurement of Key Constructs Construct Domain Indicator/Question Notes 1. Social Structure (2007) a. Socioeconomic Status • Total Household Income (annual): “What was the total income of your household over the last 12 months?” • Asset Ownership: “Does your household own any of the following: land, cars, or motorcycles?” Latent sub-factor b. Educational Opportunity • Highest Education Attended: “What is the highest level of education that you have ever attended?” • Parents’ Highest Education Attended Latent sub-factor c. Youth Community Engagement • Participation in Youth Activities: “Have you participated in or used youth activities (e.g., Karang Taruna)?” Reflects social integration d. Family Structure (at age 12) • Parents’ marital status at age 12• Lived with both biological parents at age 12 • Frequency of communication with parents Reflects early household environment 2. Cognitive Status (2014) a. Working Memory • Immediate Word Recall Task: “Please repeat all the words you can remember.” Performance-based cognitive measure b. Abstract Reasoning • Number Pattern Task: “What number is missing: 2, 4, 6, __, 10?” Performance-based cognitive measure 3. Health Risk-Taking Behaviour (2014) a. Smoking • “Have you ever chewed tobacco, smoked a pipe, smoked self-rolled cigarettes, or smoked cigarettes/cigars?” Dichotomous indicator b. Unhealthy Eating • Fast food, sweet snacks, soft drinks, fried food (in past 7 days) Items summed for total unhealthy eating score c. Physical Inactivity • “Did you do any vigorous activity for ≥ 10 minutes?” (past 7 days) (Reverse-coded) Captures lack of activity 4. Behavioural Change (2007–2014) a. Smoking Change • Still not smoking, Started smoking, Quit smoking, Still smoking Categorical transition variable b. Physical Activity Change • Remained active, Became inactive, Became active, Remained inactive Categorical transition variable Missing Data and Imputation Strategy Given the longitudinal design and sample attrition between IFLS waves, missing data presented a key analytical challenge. To preserve sample size and reduce bias, we employed Multiple Imputation by Chained Equations (MICE) using the Random Forest (RF) method in R. This approach is especially suitable for datasets involving complex, non-linear relationships between variables, as in this study’s examination of social structure, cognition, and health behaviours (Doove et al., 2014 ; Jia & Wu, 2023 ; Stekhoven & Bühlmann, 2012 ). Twenty-four of 36 key variables had missing values, and each was imputed based on all other available variables using a fully specified predictor matrix. Unique identifiers (e.g. PIDLINK) and fully observed fields were excluded. The RF method invoked via method = "rf" in the mice() function uses a flexible, non-parametric algorithm capable of handling categorical, ordinal, and continuous data. This enabled us to preserve variable-specific distributions and inter-variable relationships without relying on strong parametric assumptions (Bouhlila & Sellaouti, 2013 ). We generated five imputed datasets, each with 50 iterations per chain to ensure convergence, guided by best practices balancing computational efficiency and stability. A random seed (123) was set for reproducibility. Diagnostics confirmed convergence and consistency across imputations. To assess imputation quality, we compared key distributions pre- and post-imputation. Differences were minimal, suggesting that the imputation process preserved the plausibility and structure of the original data. All distributions remained within plausible bounds, and no implausible outliers or significant distortions were introduced. As a result, the final imputed dataset contained 11,539 complete cases across all 36 variables (0% missingness). Crucially, imputation enabled downstream structural equation modelling (SEM) that would otherwise be infeasible due to missingness. For instance, CFA and SEM could not produce valid estimates using the original dataset, whereas the imputed datasets yielded models with good fit indices (e.g., CFI and TLI), allowing robust estimation of the direct and indirect effects explored in this study. By applying MICE with RF, this study maintained analytical power, addressed potential biases from non-random missingness, and ensured that subsequent modelling of adolescent cognitive development and health behaviour trajectories reflected the full sample’s variability and complexity. Statistical Analysis This study applied Structural Equation Modelling (SEM) to examine the role of cognitive status in shaping health risk behaviours among Indonesian adolescents and young adults. Three theory-driven SEM models were estimated: Model 1: Cognitive status as a direct predictor of health risk-taking behaviour. Model 2: Cognitive status as a mediator between early-life social structure and health risk behaviour. Model 3: Cognitive status as a mediator between early-life social structure and behavioural change from adolescence to young adulthood. All latent constructs were developed based on existing theory and prior empirical work. Given the established nature of the constructs and the emphasis of this study on examining structural pathways, we focused on model-level fit and structural pathways rather than reporting standalone Confirmatory Factor Analysis (CFA) results, which aligns with practices in theory-driven SEM applications in public health. All SEM analyses were conducted using the lavaan package in R. Models controlled for age and gender, and robust standard errors were applied to account for any non-normality. Model fit was evaluated using standard SEM indices, including the Comparative Fit Index (CFI), Tucker-Lewis Index (TLI), Root Mean Square Error of Approximation (RMSEA), and Standardised Root Mean Square Residual (SRMR). Results 1. Descriptive Analysis Sample Demographics Before examining the structural model, we first describe the socio-demographic characteristics and distribution of key variables to contextualise the analysis. The study sample consisted of 11,539 Indonesian adolescents and young adults drawn from the Indonesia Family Life Survey (IFLS). At the time of outcome measurement in 2014, participants were aged 22–30 (mean = 26.1, SD = 2.4), having been 15–23 years old in 2007 (mean = 19.1, SD = 2.4) (Table 2 ). The gender distribution was 52.4% female and 47.6% male. The majority of participants (56.7%) resided in urban areas. Ethnically, 39.4% identified as Javanese, 11.9% as Sundanese, 4.9% as Batak, and the remainder from other groups. A large majority (89.6%) were Muslim, consistent with national demographics. Socio-economic indicators showed a wide range, with significant disparities between low- and high-income households. The average annual household income in 2007 was IDR 31.2 million, ranging from low-income to over IDR 2 billion. Mean land asset value was IDR 75.1 million, and vehicle assets averaged IDR 11 million. Approximately 42.4% of respondents had completed senior high school, while 10% attained higher education. Parental education levels were generally lower, with only 20.5% having completed education beyond senior high. The average distance to school was 13.25 km (SD = 45.6), highlighting infrastructural barriers in rural areas. Regarding family structure, 93% of respondents lived with both parents at age 12, and 91.4% reported parental co-residence. Parental communication, rated on a 6–30 scale, had a mean score of 14 (SD = 4.4). Youth community engagement was uneven; only 26.6% reported active participation in community activities such as Karang Taruna. Table 2 Demographic characteristics No. Variable Weighted Mean (SD) / Weighted % (95% CI) Demographics 1 Gender (% female) 6041 (52.4%) 2 Age (2007) 19.1 (2.4) 3 Age (2014) 26.1 (2.4) 4 Urban residence 6548 (56.7%) 5 Religion (% Muslim) 10,338 (89.6%) 6 Ethnicity a. Javanese 4543 (39.4%) b. Sundanese 1370 (11.9%) c. Batak 569 (4.9%) d. Others 3959 (34.0%) Socioeconomic Status (SES) 7 Total Household Income (IDR) 31.3 million (65 million) 8 Land Asset Value (IDR) 75.1 million (152 million) 9 Vehicle Asset Value (IDR) 11 million (32 million) Parental Education Level 10 Unschooled 3 (0.0%) 11 Kindergarten 723 (6.3%) 12 Elementary School 2065 (17.9%) 13 Junior High School 5602 (48.5%) 14 Senior High School 675 (5.8%) 15 D1/D2/D3 2360 (20.5%) 16 Bachelor’s Degree 106 (0.9%) 17 Master’s/PhD 5 (0.0%) Participant’s Education Level 18 Unschooled 118 (1.0%) 19 Kindergarten 1 (0.0%) 20 Elementary School 2041 (17.7%) 21 Junior High School 2830 (24.5%) 22 Senior High School 4887 (42.4%) 23 D1/D2/D3 478 (4.1%) 24 Bachelor’s Degree 1151 (10.0%) 25 Master’s/PhD 33 (0.3%) 26 Distance to School (KM) 13.25 (45.6) Family Structure (at age 12) 27 Parents Married (Yes) 10,826 (93.8%) 28 Living with Both Parents 10,548 (91.4%) 29 Frequency of Communication with Parents (1–5 scale) 4.4 (1.4) Youth Community Participation 30 Participation in Youth Activities (Yes) 3071 (26.6%) Access to Health Services 31 Health Service Access Score 3.6 (1.1) Cognitive Status Cognitive status was assessed in 2014 using two tasks: a memory recall test (scores 1–10) measuring episodic memory, and a pattern reasoning test (scores 0–7) assessing fluid intelligence. As shown in Table 3 , males scored slightly higher on memory (mean = 5.76) than females (5.56), while females had marginally higher pattern reasoning scores (2.99 vs. 2.98). Memory scores declined modestly with age, from 5.68 in ages 15–16 to 5.54 in ages 23–24, whereas pattern reasoning remained relatively stable, peaking at 3.09 in the 19–20 age group. These patterns align with established cognitive development trajectories and support the use of these measures in further modelling. Table 3 The distribution of scores for each cognitive measure, age and gender. Group N Mean Memory Score Mean Pattern Score Gender Male 6041 5.76 2.98 Female 5498 5.56 2.99 Age Group 15–16 790 5.68 2.84 16–17 1289 5.69 3.02 17–18 1348 5.71 3.00 18–19 1477 5.71 3.03 19–20 1324 5.74 3.09 20–21 1379 5.70 3.01 21–22 1395 5.63 2.91 22–23 1491 5.56 2.95 23–24 1046 5.54 2.99 Health Risk-Behaviours Health risk behaviours showed concerning patterns. In 2007, 22.3% of respondents smoked; by 2014, this rose to 36.5%, reflecting increased uptake during the transition to adulthood. Only 22.2% reported engaging in vigorous physical activity in 2014, slightly up from 19.5% in 2007. Unhealthy eating, measured on a 0–4 scale, had a mean score of 2.2 (SD = 0.8), indicating moderate consumption of unhealthy foods such as sweets, fast food, and soft drinks (Table 4 ). Table 4 Health Risk-Taking Behaviour No Variable Weighted Mean (SD) / Weighted Percentage (95% CI), N = 11,539 1. Smoking 2014 (yes) 4208 (36.5%) 2 Smoking 2007 (yes) 2573 (22.3%) 3 Physical activity 2014 (yes) 2563 (22.2%) 4 Physical activity 2007 (yes) 2249 (19.5%) 5 Unhealthy eating behaviour (scale 1–4) 2.2 (0.8) The descriptive analysis reveals important behavioural patterns during the transition from adolescence to young adulthood. For smoking, the majority of respondents remained non-smokers across both waves (62.9%), which is encouraging. However, a substantial proportion either continued smoking (21.7%) or initiated smoking during the period (14.8%) (Table 5 ). Alarmingly, very few adolescents reported quitting smoking (0.6%), highlighting persistently low cessation rates. This finding is particularly concerning, as it underscores how difficult it is to quit once smoking behaviour has been established. Table 5 Frequency of behaviour changes No. Status Category Smoking Status Frequency (Smoking) Physical Activity Status Frequency (Physical Activity) 1 Still not engaging 0 = Still Not Smoking 7,258 (62.9) 0 = Still Active 7,615 (66) 2 Positive change 1 = Quit Smoking 73 (0.6) 1 = Started Active 1,675 (14.5) 3 Negative change 2 = Started Smoking 1,708 (14.8) 2 = Quit Active 1,361 (11.8) 4 Still engaging 3 = Still Smoking 2,500 (21.7%) 3 = Still Not Active 888 (7.7) Total — 11,539 — 11,539 2. Structural Models Model 1: Cognitive status and health Risk Behaviour Model 2: Cognitive Status as a Mediator between Social Structure and Health Risk Behaviour Model 3: Cognitive Status as a Mediator between Social Structure and Behavioural Change Discussion This study examined the role of cognitive status as both a direct predictor and a mediator of health risk-taking behaviour among adolescents and young adults in Indonesia using SEM across three models. The findings highlight a developmental pathway where cognitive function partially explains how early social inequalities lead to healthier behavioural outcomes, offering new insight into the mechanisms of adolescent behavioural development in a developing country context marked by health inequalities and a rapid second demographic transition characterised by falling fertility and an increasing aging population. The descriptive data reveal considerable heterogeneity in the sample’s socioeconomic and demographic background. The average annual household income in 2007 was IDR 31.2 million, ranging widely from low-income households to over IDR 2 billion, with substantial variability in land and vehicle assets. Educational attainment showed that 42.4% of respondents completed senior high school and 10% attained higher education, whereas parental education levels were generally lower, indicating intergenerational disparities. The average distance to school was notably high at 13.25 km (SD = 45.6), reflecting infrastructural and geographic barriers prevalent in Indonesia’s rural and remote areas (World Bank, 2017 ). Family structures were predominantly stable, with over 90% living with both parents at age 12 and moderate parental communication levels. However, youth community engagement was limited, with only 26.6% participating in activities such as Karang Taruna. This socio-structural context frames the latent social structure construct and likely shapes cognitive development and health behaviour patterns, as supported by previous studies emphasizing the role of structural conditions in adolescent development (Bradley & Corwyn, 2002 ; Evans & Kim, 2013 ). First, the analysis confirmed that cognitive status measured through memory recall and abstract reasoning was a statistically significant protective factor against risk behaviours such as smoking, unhealthy eating, and physical inactivity. These findings are in line with prior research in high-income contexts, which suggests that cognitive skills underpin health literacy, future planning, and self-regulation, all of which are essential to avoid risky health behaviours (Batty et al., 2007 ; Hackman et al., 2010 ; Reyna & Farley, 2006 ). Similarly, studies from other developing settings also emphasize how self-regulatory strategies often shaped by social support systems can mediate risk behaviours in youth. For example, Gaspar de Matos et al., ( 2016 ) found that children and adolescents with stronger self-regulation and social support had greater healthy eating awareness, highlighting the importance of individual cognitive capacities reinforced by environmental conditions. Second, cognitive status emerged as a meaningful mediator in the relationship between early-life social structure and later health risk behaviours. Adolescents embedded in structurally advantaged environments characterised by higher parental education, household income, family stability, and community participation developed stronger cognitive capacities by young adulthood. These capacities, in turn, reduced their engagement in health risk behaviours. Although the mediation was partial, the pathway is consistent with life-course and developmental models positing cognition as a key mechanism linking structural conditions to behavioural health outcomes (Ferrer & McArdle, 2010 ; Halfon & Hochstein, 2002 ; Shonkoff et al., 2009 ). However, it is important to note that the variance explained in cognitive status (R² = 0.007) was relatively low, indicating that factors beyond those included in the structural model likely shape adolescent cognitive development. Early-life nutrition, psychosocial stimulation, and exposure to environmental stressors may play a crucial role in shaping cognitive outcomes (Grantham-McGregor et al., 2007 ; Noble et al., 2005 ). This reinforces the importance of holistic interventions that integrate health, nutrition, and educational support during childhood and adolescence. This perspective aligns with qualitative research from the Democratic Republic of Congo, which underscores how poverty, peer influence, and parental behaviour shape youth involvement in risky behaviours such as alcohol use and violence often through reduced self-regulatory and social protective factors (Kohli et al., 2018 ). Third, the comparison of the three structural models provides important insights into how cognitive status operates in different behavioural contexts. Model 1 demonstrated that cognition had a direct protective effect on static health risk behaviours in young adulthood. In Model 2, cognition functioned as a partial mediator, linking early social advantage to reduced risk behaviours in 2014. However, in Model 3 focused on behavioural change over time cognition’s effect was mainly indirect, while early social structure had a stronger total effect on behavioural transitions. This highlights a critical distinction: cognitive skills may play a more direct role in supporting immediate health decisions (e.g., avoiding smoking), but structural factors are more powerful in influencing long-term behaviour patterns (e.g., quitting smoking or maintaining physical activity). The contrast between Model 2 and Model 3 is especially relevant. In Model 2, cognitive function was significantly associated with contemporaneous health risk behaviours, even after accounting for social structure, suggesting its role in enabling informed decision-making and behavioural restraint. In contrast, Model 3 showed that cognitive function played an indirect role, mediating the influence of early social advantage on changes in health behaviours across time. The stronger total effect of social structure in Model 3 suggests that long-term behavioural shifts are more deeply shaped by socioeconomic conditions than by cognitive capacity alone. This distinction adds depth to our understanding of how and when cognitive status matters most, highlighting that cognitive skills are critical for sustaining good behaviour but less influential when individuals are undergoing major behavioural transitions influenced by external contexts. These findings align with Bourdieu’s theory of habitus, where early-life structural positions shape durable dispositions and behavioural routines (Bourdieu, 1977 ). This may help explain why structural conditions had a stronger effect on behaviour change: once established, behaviours become embedded in social environments that are difficult to shift through individual cognitive effort alone. The results also provide empirical support for the cumulative disadvantage framework, which posits that early structural inequalities compound over time to produce long-term disparities in health outcomes (Dannefer, 2003 ). Gender-specific pathways also emerged in the analysis, with males significantly more likely to engage in risky health behaviours. This reflects gendered norms in Indonesia that normalise smoking among boys while discouraging it among girls (Fithria et al., 2021 ; Kodriati et al., 2020 ). Such patterns suggest the need for gender-sensitive health interventions that address social norms alongside individual cognition. This layered comparison across models strengthens the originality of the study and demonstrates the utility of using SEM to parse these complex developmental and structural interactions in youth health research. Such modelling approaches remain rare in studies of adolescents in developing countries like Indonesia, where demographic shifts and structural inequalities co-exist. This study contributes to a more nuanced life-course understanding of adolescent health. From a policy perspective, the findings reinforce the value of integrated interventions. Programmes that invest in early-life structural conditions such as improving access to education, reducing poverty, promoting family cohesion, and enhancing community engagement can yield dual dividends: they promote cognitive development and reduce long-term health risks (Bradley & Corwyn, 2002 ; Marmot et al., 2008 ). While targeted cognitive development initiatives such as programs aimed at improving memory, abstract reasoning, and decision-making may offer valuable support for youth health behaviours, they should not be pursued in isolation. Without addressing broader structural inequalities, such interventions risk disproportionately benefiting more advantaged adolescents, potentially exacerbating existing disparities. For instance, schools in low-resource settings may lack the infrastructure to implement cognitive enhancement programs, and young people in remote or underserved areas may have limited access to such opportunities. Research shows that socioeconomic disadvantage is closely associated with lower performance in memory and reasoning tasks, which may reduce the effectiveness of cognitive interventions unless underlying conditions are also addressed (Noble et al., 2005 ; von Stumm, 2017 ). Therefore, cognitive programming must be implemented alongside structural reforms such as investments in equitable education, nutrition, and health services to ensure access, effectiveness, and cultural relevance across diverse settings (Blair & Raver, 2016 ; OECD, 2025 ). In Indonesia, these findings hold particular relevance given the country's ongoing demographic bonus, a growing young population and persistent socioeconomic disparities in health and education access. Scalable strategies could include embedding cognitive training in school curricula, supporting youth engagement platforms like Karang Taruna, and expanding adolescent access to preventive health and educational services. However, these efforts must be tailored to the local sociocultural context. For instance, given gender norms that stigmatise smoking among girls but normalise it for boys, behaviour change strategies should account for such social pressures. Furthermore, rural-urban disparities in health access and education may limit the reach of interventions unless accompanied by broader structural reforms. Addressing these inequalities is essential for achieving equitable outcomes during this critical demographic transition. These results offer timely and relevant evidence for adolescent health policy in Indonesia and similar developing countries. They affirm the role of cognitive development as a cross-cutting determinant of health and illustrate how early social context sets the foundation for youth health trajectories. The integration of cognitive and structural interventions could represent a cost-effective, life-course approach to reducing behavioural health risks among the next generation. Nevertheless, several limitations must be noted. First, although the study employs a longitudinal design, unmeasured confounding factors such as personality traits or mental health status could bias the relationships observed. Second, the cognitive measures used for memory and pattern recognition do not capture the full range of executive functions that may influence health behaviour. Third, the indicators of health risk behaviour were limited to smoking, diet, and physical activity, omitting other relevant behaviours such as alcohol use or sexual risk. Finally, the behavioural change variable was constructed from self-reported ordinal data, which may be subject to recall bias or social desirability effects. Future research should consider incorporating a broader array of cognitive, behavioural, and psychosocial indicators to better capture the complexity of adolescent development. It would also be valuable to examine how specific community-level factors, such as peer norms, school quality, or local health services, moderate the cognitive-behaviour relationship. Mixed-methods approaches could enrich understanding of how adolescents interpret and act on health risks in structurally diverse settings. Ultimately, these refinements could enhance the effectiveness and equity of adolescent health interventions in Indonesia and other developing settings. Conclusion This study contributes new evidence to understanding the role of cognitive development in adolescent health behaviour within a developing country context. Using three SEM models, it establishes that cognitive status both directly and indirectly shapes risk-taking behaviour, and partially explains the impact of early social advantage on later outcomes. Importantly, the study highlights that cognition has a more pronounced effect on contemporaneous health decisions, while long-term behavioural changes are more strongly influenced by structural conditions. These findings support the need for a dual-focus strategy that combines cognitive development efforts with broader structural reforms. In the Indonesian context marked by demographic opportunity but persistent inequalities such an integrated approach could help support healthier youth transitions. However, to ensure these interventions do not inadvertently widen disparities, they must be equitably designed, resourced, and sensitive to the sociocultural and geographic diversity of the population. Declarations Ethical approval statement This study used de-identified secondary data from the Indonesia Family Life Survey (IFLS), which received ethical approval from the Institutional Review Boards at RAND and the University of Gadjah Mada. Additional ethical clearance for this analysis was obtained through the University of Sheffield’s ethics application system (Reference Number: 057057). All procedures were conducted in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards. Funding No funding was received for conducting this study . Acknowledgement This research was supported by The Indonesian Education Scholarship Program (Beasiswa Pendidikan Indonesia - BPI), funded by the Centre for Higher Education Funding and Assessment, Ministry of Higher Education, Science, and Technology of the Republic of Indonesia. The author gratefully acknowledges this support. Disclosure statement No potential conflict interest was reported by the authors. Data availability statement The datasets used in this study are publicly available in the Indonesia Family Life Survey (IFLS) at https://www.rand.org/well-being/social-and-behavioral-policy/data/FLS/IFLS/ifls5.html. The dataset analysed during the current study is available from correspondence author on reasonable request. Clinical trial registration Clinical trial registration: Not applicable References Arnett, J. J. (2000). Emerging adulthood: A theory of development from the late teens through the twenties. American Psychologist , 55 (5), 469–480. https://doi.org/10.1037/0003-066X.55.5.469 Batty, G. D., Deary, I. J., & Gottfredson, L. S. (2007). Premorbid (early life) IQ and later mortality risk: Systematic review. Annals of Epidemiology , 17 (4), 278–288. https://doi.org/10.1016/j.annepidem.2006.07.010 Blair, C., & Raver, C. C. (2016). Poverty, Stress, and Brain Development: New Directions for Prevention and Intervention. Academic Pediatrics , 16 (3 Suppl), S30-36. https://doi.org/10.1016/j.acap.2016.01.010 Bouhlila, D. S., & Sellaouti, F. (2013). Multiple imputation using chained equations for missing data in TIMSS: A case study. Large-Scale Assessments in Education , 1 (1), 4. https://doi.org/10.1186/2196-0739-1-4 Bourdieu, P. (with Internet Archive). (1977). Outline of a theory of practice . Cambridge ; New York : Cambridge University Press. http://archive.org/details/outlineoftheoryo0000bour Bradley, R. H., & Corwyn, R. F. (2002). Socioeconomic status and child development. Annual Review of Psychology , 53 (1), 371–399. https://doi.org/10.1146/annurev.psych.53.100901.135233 Brieant, A., Peviani, K. M., Lee, J. E., King-Casas, B., & Kim-Spoon, J. (2021). Socioeconomic Risk for Adolescent Cognitive Control and Emerging Risk-Taking Behaviors. Journal of Research on Adolescence , 31 (1), 71–84. https://doi.org/10.1111/jora.12583 Clark, C., Rodgers, B., Caldwell, T., Power, C., & Stansfeld, S. (2007). Childhood and adulthood psychological ill health as predictors of midlife affective and anxiety disorders: The 1958 British Birth Cohort. Archives of General Psychiatry , 64 (6), 668–678. https://doi.org/10.1001/archpsyc.64.6.668 Costa, P. T., McCrae, R. R., & Dye, D. A. (1991). Facet Scales for Agreeableness and Conscientiousness: A Revision of the NEO Personality Inventory. Personality and Individual Differences , 12 (9), 887–898. https://doi.org/10.1016/0191-8869(91)90177-D Daly, A. N., O’Sullivan, E. J., & Kearney, J. M. (2022). Considerations for health and food choice in adolescents. Proceedings of the Nutrition Society , 81 (1), 75–86. https://doi.org/10.1017/S0029665121003827 Dannefer, D. (2003). Cumulative advantage/disadvantage and the life course: Cross-fertilizing age and social science theory. The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences , 58 (6), S327-337. https://doi.org/10.1093/geronb/58.6.s327 Doove, L. L., Van Buuren, S., & Dusseldorp, E. (2014). Recursive partitioning for missing data imputation in the presence of interaction effects. Computational Statistics & Data Analysis , 72 , 92–104. https://doi.org/10.1016/j.csda.2013.10.025 Evans, G. W., & Kim, P. (2013). Childhood poverty, chronic stress, self‐regulation, and coping. Child Development Perspectives , 7 (1), 43–48. https://doi.org/10.1111/cdep.12013 Ferrer, E., & McArdle, J. J. (2010). Longitudinal modeling of developmental changes in psychological research. Current Directions in Psychological Science , 19 (3), 149–154. https://doi.org/10.1177/0963721410370300 Fithria, F., Adlim, M., Jannah, S. R., & Tahlil, T. (2021). Indonesian adolescents’ perspectives on smoking habits: A qualitative study. BMC Public Health , 21 (1), 82. https://doi.org/10.1186/s12889-020-10090-z Gaspar de Matos, M., Palmeira, A. L., Gaspar, T., De Wit, J. B. F., & Luszczynska, A. (2016). Social support influences on eating awareness in children and adolescents: The mediating effect of self-regulatory strategies. Global Public Health , 11 (4), 437–448. https://doi.org/10.1080/17441692.2015.1094106 Grantham-McGregor, S., Cheung, Y. B., Cueto, S., Glewwe, P., Richter, L., Strupp, B., & International Child Development Steering Group. (2007). Developmental potential in the first 5 years for children in developing countries. Lancet (London, England) , 369 (9555), 60–70. https://doi.org/10.1016/S0140-6736(07)60032-4 Hackman, D. A., Farah, M. J., & Meaney, M. J. (2010). Socioeconomic status and the brain: Mechanistic insights from human and animal research. Nature Reviews Neuroscience , 11 (9), 651–659. Halfon, N., & Hochstein, M. (2002). Life course health development: An integrated framework for developing health, policy, and research. The Milbank Quarterly , 80 (3), 433–479, iii. https://doi.org/10.1111/1468-0009.00019 Jia, F., & Wu, W. (2023). A comparison of multiple imputation strategies to deal with missing nonnormal data in structural equation modeling. Behavior Research Methods , 55 (6), 3100–3119. https://doi.org/10.3758/s13428-022-01936-y Junger, M., & van Kampen, M. (2010). Cognitive ability and self-control in relation to dietary habits, physical activity and bodyweight in adolescents. The International Journal of Behavioral Nutrition and Physical Activity , 7 , 22. https://doi.org/10.1186/1479-5868-7-22 Kodriati, N., Hayati, E. N., Santosa, A., & Pursell, L. (2020). Perceived social benefits versus perceived harms of smoking among Indonesian boys aged 12–16 years: A secondary analysis of Global Youth Tobacco Survey 2014. Tobacco Prevention & Cessation , 6 , 8. https://doi.org/10.18332/tpc/115034 Kohli, A., Remy, M. M., Binkurhorhwa, A. K., Mitima, C. M., Mirindi, A. B., Mwinja, N. B., Banyewesize, J. H., Ntakwinja, G. M., Perrin, N. A., & Glass, N. (2018). Preventing risky behaviours among young adolescents in eastern Democratic Republic of Congo: A qualitative study. Global Public Health , 13 (9), 1241–1253. https://doi.org/10.1080/17441692.2017.1317009 Marmot, M., Friel, S., Bell, R., Houweling, T. A. J., Taylor, S., & Commission on Social Determinants of Health. (2008). Closing the gap in a generation: Health equity through action on the social determinants of health. Lancet (London, England) , 372 (9650), 1661–1669. https://doi.org/10.1016/S0140-6736(08)61690-6 Ng, M., Freeman, M. K., Fleming, T. D., Robinson, M., Dwyer-Lindgren, L., Thomson, B., Wollum, A., Sanman, E., Wulf, S., Lopez, A. D., Murray, C. J. L., & Gakidou, E. (2014). Smoking prevalence and cigarette consumption in 187 countries, 1980-2012. JAMA , 311 (2), 183–192. https://doi.org/10.1001/jama.2013.284692 Noble, K. G., Norman, M. F., & Farah, M. J. (2005). Neurocognitive correlates of socioeconomic status in kindergarten children. Developmental Science , 8 (1), 74–87. https://doi.org/10.1111/j.1467-7687.2005.00394.x OECD. (2025). Reducing Inequalities by Investing in Early Childhood Education and Care: Project methodology for translating research into early childhood education and care policies . OECD. https://www.oecd.org/en/publications/reducing-inequalities-by-investing-in-early-childhood-education-and-care_b78f8b25-en/full-report/project-methodology-for-translating-research-into-early-childhood-education-and-care-policies_93f1eb79.html Patton, G. C., Sawyer, S. M., Santelli, J. S., Ross, D. A., Afifi, R., Allen, N. B., Arora, M., Azzopardi, P., Baldwin, W., Bonell, C., Kakuma, R., Kennedy, E., Mahon, J., McGovern, T., Mokdad, A. H., Patel, V., Petroni, S., Reavley, N., Taiwo, K., … Viner, R. M. (2016). Our future: A Lancet commission on adolescent health and wellbeing. The Lancet , 387 (10036), 2423–2478. https://doi.org/10.1016/S0140-6736(16)00579-1 Reyna, V. F., & Farley, F. (2006). Risk and rationality in adolescent decision making: Implications for theory, practice, and public policy. Psychological Science in the Public Interest , 7 (1), 1–44. Septiono, W., Kuipers, M. A. G., Ng, N., & Kunst, A. E. (2022). Self-reported exposure of Indonesian adolescents to online and offline tobacco advertising, promotion and sponsorship (TAPS). Tobacco Control , 31 (1), 98–105. https://doi.org/10.1136/tobaccocontrol-2020-056080 Shonkoff, J. P., Boyce, W. T., & McEwen, B. S. (2009). Neuroscience, molecular biology, and the childhood roots of health disparities: Building a new framework for health promotion and disease prevention. JAMA , 301 (21), 2252–2259. https://doi.org/10.1001/jama.2009.754 Steinberg, L. (2008). A social neuroscience perspective on adolescent risk-taking. Developmental Review , 28 (1), 78–106. https://doi.org/10.1016/j.dr.2007.08.002 Stekhoven, D. J., & Bühlmann, P. (2012). MissForest—Nonparametric missing value imputation for mixed-type data. Bioinformatics , 28 (1), 112–118. https://doi.org/10.1093/bioinformatics/btr597 Strauss, J., Witoelar, F., & Sikoki, B. (2016). The Fifth Wave of the Indonesia Family Life Survey: Overview and Field Report: Volume 1 . RAND Corporation. https://doi.org/10.7249/WR1143.1 UNFPA. (2025). World Population Dashboard -Indonesia | United Nations Population Fund . https://www.unfpa.org/data/world-population/ID Viner, R. M., Ozer, E. M., Denny, S., Marmot, M., Resnick, M., Fatusi, A., & Currie, C. (2012). Adolescence and the social determinants of health. Lancet (London, England) , 379 (9826), 1641–1652. https://doi.org/10.1016/S0140-6736(12)60149-4 von Stumm, S. (2017). Socioeconomic status amplifies the achievement gap throughout compulsory education independent of intelligence. Intelligence , 60 , 57–62. https://doi.org/10.1016/j.intell.2016.11.006 Whalley, L. J., & Deary, I. J. (2001). Longitudinal cohort study of childhood IQ and survival up to age 76 . https://doi.org/10.1136/bmj.322.7290.819 World Bank. (2017). Improving Education Quality in Indonesia’s Poor Rural and Remote Areas [Text/HTML]. World Bank. https://www.worldbank.org/en/results/2017/12/22/improving-education-quality-in-indonesia-poor-rural-and-remote-areas Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Reviews received at journal 12 Sep, 2025 Reviews received at journal 10 Sep, 2025 Reviewers agreed at journal 20 Aug, 2025 Reviewers agreed at journal 13 Aug, 2025 Reviewers invited by journal 11 Aug, 2025 Editor invited by journal 01 Aug, 2025 Editor assigned by journal 31 Jul, 2025 Submission checks completed at journal 31 Jul, 2025 First submitted to journal 29 Jul, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7244173","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":502187305,"identity":"92c1ddc8-b5a6-496b-abcb-eb4e785ae41f","order_by":0,"name":"Gisely Vionalita","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABAElEQVRIiWNgGAWjYDACZiB+wGAD58uAyQQ2AloSGNLgfB7CWsAKGA6jaWHAo0Xenf3hg8S283Ly7WfMHn75dZiHgf3wA4YHZbi1GB7mMTZIbLttbHAmx9xYtg+ohSfNgCHhHB4tzTxsEkAtiRsYcsykJXuAWhhyGBgS2/BpYX/+I7HtXP38/jdQLfxv8GuRZ2YwAyo4kMBwI8dM8sMPoBYJArYYMPMYSyScSzbccONZmTRjQzrQnc8MDuDzi3z/8YcfPpTZycv3J2+T/PHHWo6fP/nhwx94QszgAJBghMYCM28bJEYO4NYAtKUBRP6BcBh//MGndhSMglEwCkYqAABzZE7R6kkikwAAAABJRU5ErkJggg==","orcid":"","institution":"University of Sheffield","correspondingAuthor":true,"prefix":"","firstName":"Gisely","middleName":"","lastName":"Vionalita","suffix":""},{"id":502187307,"identity":"a08bb612-142d-4b90-a145-264811e913d8","order_by":1,"name":"Nathan Hughes","email":"","orcid":"","institution":"University of Sheffield","correspondingAuthor":false,"prefix":"","firstName":"Nathan","middleName":"","lastName":"Hughes","suffix":""},{"id":502187309,"identity":"56a83986-00b5-4089-9a18-8d403db709c3","order_by":2,"name":"Alvaro Martinez-Perez","email":"","orcid":"","institution":"University of Sheffield","correspondingAuthor":false,"prefix":"","firstName":"Alvaro","middleName":"","lastName":"Martinez-Perez","suffix":""}],"badges":[],"createdAt":"2025-07-29 14:08:32","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7244173/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7244173/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":89388823,"identity":"5f2a4c67-5b0c-472d-9509-9c77031fd667","added_by":"auto","created_at":"2025-08-19 12:44:01","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":35786,"visible":true,"origin":"","legend":"\u003cp\u003eStructural Model of Direct Relationship between Cognitive Status and Health Risk Behaviour (Model 1)\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eModel 1\u003c/em\u003e: Cognitive status in 2014 had a significant negative effect on health risk behaviour (standardised β = -0.12, p \u0026lt; 0.001) (Figure 1). While modest in magnitude, the effect size suggests a meaningful protective role of cognitive function in shaping behavioural choices. This suggests that adolescents and young adults with higher cognitive functioning were less likely to engage in smoking, unhealthy eating, and physical inactivity by early adulthood. This protective role of cognitive ability remained statistically robust even after adjusting for gender and age. Gender was a strong predictor of health risk behaviour. Males were significantly more likely to report engaging in risky health behaviours (standardised β = 0.90, p \u0026lt; 0.001), particularly smoking, consistent with entrenched gender norms around tobacco use in Indonesia. Age also showed a modest but significant positive association with health risk behaviour (β = 0.04, p \u0026lt; 0.001), indicating a gradual accumulation of risk exposures as individuals transitioned from adolescence to young adulthood. The model 1 fit was strong (CFI = 0.975, TLI = 0.963, RMSEA = 0.038, SRMR = 0.035), confirming that the hypothesised structure provided a good representation of the observed data. Cognitive status itself was moderately explained by memory and pattern task performance (R² = 0.007), though variance explained was relatively low, likely due to limited cognitive measures.\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7244173/v1/5f73e66d5831ad7fe0e0f98b.jpg"},{"id":89386509,"identity":"adc19ee1-5a2c-4f70-a3a1-89cc51e22a2f","added_by":"auto","created_at":"2025-08-19 12:36:01","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":54787,"visible":true,"origin":"","legend":"\u003cp\u003eStructural Model of Cognitive Status and Mediate role for Health Risk Behaviour (Model 2)\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eModel 2\u003c/em\u003e: This model (Figure 2) examined the role of cognitive status as a mediator linking early social structure in 2007 to health risk behaviours in 2014. The findings show that social structure had a significant direct negative effect on health risk behaviour (standardised β = -0.13, p \u0026lt; 0.001), suggesting that adolescents with more advantaged social conditions were less likely to engage in smoking, physical inactivity, or unhealthy eating by early adulthood. Social structure also significantly predicted cognitive status (β = 0.72, p \u0026lt; 0.001). Although the direct path from cognition to risk behaviour was not statistically significant (β = -0.03, p = 0.138), the overall indirect effect via cognitive status remained significant (β = -0.03, p = 0.138), the overall indirect effect of social structure on risk behaviour through cognition was significant (standardised β_indirect = -0.09, p \u0026lt; 0.001). This supports partial mediation, where cognitive function serves as a conduit through which social advantage reduces the likelihood of risky health behaviours. The total effect of social structure on health risk behaviour was β_total = -0.12, indicating that cognition accounted for approximately 30–35% of the total effect. These findings reinforce the theoretical argument that early-life structural advantage contributes to healthier trajectories partly by fostering cognitive capacity, which in turn shapes behavioural self-regulation. The model demonstrated a strong fit (CFI = 0.963, TLI = 0.953, RMSEA = 0.033, SRMR = 0.039), further validating the robustness of these relationships in a large national sample.\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7244173/v1/347a3d64cbccd41b58b6fc7e.jpg"},{"id":89386500,"identity":"055c072c-b553-4f4a-b292-81e269179b1a","added_by":"auto","created_at":"2025-08-19 12:36:01","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":62939,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStructural Model of Cognitive Status as Mediate Role for Behaviour Changes (Model 3)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eModel 3\u003c/em\u003e: This model investigates how early-life social structure affects behavioural changes in smoking and physical activity from adolescence to young adulthood. The standardized path from social structure to behavioural change was significant and negative (β = -0.31), suggesting that adolescents with higher social advantage were less likely to experience negative behaviour transitions by 2014. Cognitive status was strongly predicted by social structure (β = 0.72), indicating that early social conditions were associated with improved cognitive function in young adulthood (Figure 3). However, the direct effect of cognitive status on behavioural change was small and not statistically significant (β = 0.05, p = 0.257), suggesting that cognition did not independently predict changes in smoking or physical activity over time. Despite the non-significant direct path, the indirect effect of social structure on behavioural change through cognitive status was statistically significant (standardised β_indirect = -0.22, p \u0026lt; 0.001), supporting the role of cognition as a partial mediator. This indicates that cognitive improvements linked to early social advantage contributed meaningfully to later behavioural outcomes. The total effect of social structure on behavioural change (standardised β_total = -0.17) reflects both direct and mediated influences, confirming that structural conditions during adolescence exert long-term impacts on health-related behaviour patterns. The model demonstrated acceptable fit to the data (CFI = 0.934, TLI = 0.915, RMSEA = 0.040, SRMR = 0.037), supporting the robustness of these findings.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-7244173/v1/fa57dc968324a6b82d376e6f.jpg"},{"id":89390855,"identity":"22cb2ba1-bf96-4628-b7d9-7ce7899a3e19","added_by":"auto","created_at":"2025-08-19 13:00:02","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1370682,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7244173/v1/5e628540-c8d2-4eea-aa81-0e4fc26f6adf.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Understanding Youth Health Risk Behaviours in Indonesia through a Structural Model of Social and Cognitive Determinants: A Longitudinal Analysis","fulltext":[{"header":"Introduction","content":"\u003cp\u003eAdolescence and young adulthood represent a critical developmental life stage during which individuals undergo significant cognitive maturation while also becoming increasingly exposed to health risk behaviours, including smoking, physical inactivity, and unhealthy dietary patterns. These years are marked by both vulnerability and opportunity: cognitive development including memory, abstract reasoning, and executive function enables adolescents to make complex decisions, anticipate long-term consequences, and regulate impulses (Junger \u0026amp; van Kampen, \u003cspan class=\"CitationRef\"\u003e2010\u003c/span\u003e; Whalley \u0026amp; Deary, \u003cspan class=\"CitationRef\"\u003e2001\u003c/span\u003e). At the same time, adolescents are influenced by shifting social environments, including peer pressure, changing family dynamics, and community expectations, which shape behavioural patterns that often persist into adulthood (Arnett, \u003cspan class=\"CitationRef\"\u003e2000\u003c/span\u003e; Steinberg, \u003cspan class=\"CitationRef\"\u003e2008\u003c/span\u003e). In high-income countries, a growing body of evidence has shown that adolescents with stronger cognitive function are less likely to engage in health risk behaviours such as smoking, substance use, and physical inactivity (Batty et al., \u003cspan class=\"CitationRef\"\u003e2007\u003c/span\u003e; Clark et al., \u003cspan class=\"CitationRef\"\u003e2007\u003c/span\u003e; Daly et al., \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e; Hackman et al., \u003cspan class=\"CitationRef\"\u003e2010\u003c/span\u003e). Cognitive abilities particularly those related to memory and reasoning have been linked to improved decision-making and better health-related judgement (Reyna \u0026amp; Farley, \u003cspan class=\"CitationRef\"\u003e2006\u003c/span\u003e). However, while this evidence base is well developed in high-income contexts, far less is known about how these relationships manifest in upper-middle-income countries, where social and economic transitions, regional disparities, and structural inequalities may differently shape both cognitive development and health behaviour trajectories (Patton et al., \u003cspan class=\"CitationRef\"\u003e2016\u003c/span\u003e; Viner et al., \u003cspan class=\"CitationRef\"\u003e2012\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eAlthough Indonesia is now classified as an upper-middle-income country, it continues to face uneven development, particularly in access to education and healthcare services. Youth in Indonesia, who comprise approximately 24% of the national population (aged 10\u0026ndash;24) (UNFPA, \u003cspan class=\"CitationRef\"\u003e2025\u003c/span\u003e), navigate multiple layers of socio‑cultural and economic transitions that are likely to influence both their cognitive and behavioural outcomes. In this context, understanding the pathways through which cognitive development mediates the relationship between early social advantage and health behaviours is essential for designing effective public health interventions. The role of cognitive status in shaping adolescent health behaviour is hypothesised to be twofold. First, cognitive skills may exert a direct protective effect, enabling individuals to evaluate health risks more effectively and make informed behavioural choices (Costa et al., \u003cspan class=\"CitationRef\"\u003e1991\u003c/span\u003e; Reyna \u0026amp; Farley, \u003cspan class=\"CitationRef\"\u003e2006\u003c/span\u003e). Second, cognitive development may act as a mediating mechanism through which early-life structural conditions such as stable family conditions, higher parental education, and community support translate into healthier adolescent behaviour (Brieant et al., \u003cspan class=\"CitationRef\"\u003e2021\u003c/span\u003e; Hackman et al., \u003cspan class=\"CitationRef\"\u003e2010\u003c/span\u003e). Life-course theory suggests that the structural conditions experienced during adolescence leave lasting imprints on developmental outcomes, including cognitive capacity, which in turn influence future health trajectories (Halfon \u0026amp; Hochstein, \u003cspan class=\"CitationRef\"\u003e2002\u003c/span\u003e; Shonkoff et al., \u003cspan class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e\n\u003cp\u003eThis framework is particularly relevant in the Indonesian context. Evidence shows that Indonesian adolescent males have among the highest smoking prevalence rates globally (Ng et al., \u003cspan class=\"CitationRef\"\u003e2014\u003c/span\u003e), while adolescent girls face gendered barriers in physical activity and food autonomy (Septiono et al., \u003cspan class=\"CitationRef\"\u003e2022\u003c/span\u003e). At the same time, disparities in family stability, education quality, and access to youth support services are likely to contribute to wide variability in both cognitive and behavioural outcomes across socioeconomic groups. Despite these dynamics, there is a notable lack of empirical research in upper-middle-income countries on how cognitive function interacts with structural factors to shape health behaviours during the transition from adolescence to adulthood.\u003c/p\u003e\n\u003cp\u003eThis study addresses that gap by analysing longitudinal Indonesian survey data to test a conceptual model in which cognitive status functions both as a direct predictor and a mediator of health risk behaviours. Drawing on life-course theory and the social determinants of health framework, this study uses structural equation modelling (SEM) to examine how adolescent cognitive function measured through verbal memory and fluid reasoning interacts with early social inequalities to shape health risk behaviours and behavioural changes by early adulthood. Specifically, the study addresses two research questions; (1) what is the association between cognitive status and health risk-taking behaviour among adolescents and young adults in Indonesia? And (2) does cognitive status mediate the relationship between early-life social structure and health risk-taking behaviour? The findings will seek to inform youth-focused public health strategies by providing evidence on how cognitive development might serve as a protective mechanism in reducing risk behaviours, particularly in rapidly developing contexts such as Indonesia.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cb\u003eStudy Design and Data Source\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe aim of this study was to investigate the direct and indirect effects of cognitive status on health risk-taking behaviour among Indonesian adolescents and young adults, using longitudinal structural equation modelling. This study used data from the Indonesia Family Life Survey (IFLS) a longitudinal, nationally representative household survey spanning five waves from 1993 to 2014. For this analysis, we used data from Wave 4 (2007) and Wave 5 (2014) to examine the relationships between early-life social structure, cognitive function, and health risk-taking behaviours in adolescence and young adulthood. The IFLS employed a multistage stratified sampling method and included 13 provinces representing approximately 83% of the Indonesian population. Respondents aged 15–22 in 2007 were followed up as young adults (aged 23–30) in 2014. The survey included a wide array of health, economic, cognitive, and psychosocial measures collected via interviewer-administered and self-completed questionnaires (Strauss et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2016\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cb\u003eSample Selection\u003c/b\u003e\u003c/p\u003e\u003cp\u003eWe restricted the sample to individuals aged 15–22 years in 2007 and 23–30 years in 2014, yielding a total of 11,539 eligible participants. A balanced panel dataset was created by retaining respondents who participated in both waves, resulting in a final analytical sample of 2,525 individuals. While this resulted in attrition (~ 57%), unbalanced panel data were retained for supplementary and robustness analyses to enhance generalisability.\u003c/p\u003e\u003cp\u003e\u003cb\u003eMeasurement of Key Constructs\u003c/b\u003e\u003c/p\u003e\u003cp\u003eAll variables were derived from the Indonesia Family Life Survey (IFLS) questionnaire and mapped to key theoretical constructs. Latent constructs were defined using validated survey items consistent with prior studies and grounded in the life-course framework (see Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eMeasurement of Key Constructs\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"4\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eConstruct\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDomain\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eIndicator/Question\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eNotes\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003e1. Social Structure (2007)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ea. Socioeconomic Status\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e• Total Household Income (annual): “What was the total income of your household over the last 12 months?”\u003c/p\u003e\u003cp\u003e• Asset Ownership: “Does your household own any of the following: land, cars, or motorcycles?”\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLatent sub-factor\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eb. Educational Opportunity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e• Highest Education Attended: “What is the highest level of education that you have ever attended?”\u003c/p\u003e\u003cp\u003e• Parents’ Highest Education Attended\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eLatent sub-factor\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ec. Youth Community Engagement\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e• Participation in Youth Activities: “Have you participated in or used youth activities (e.g., Karang Taruna)?”\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eReflects social integration\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ed. Family Structure (at age 12)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e• Parents’ marital status at age 12• Lived with both biological parents at age 12\u003c/p\u003e\u003cp\u003e• Frequency of communication with parents\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eReflects early household environment\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003e2. Cognitive Status (2014)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ea. Working Memory\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e• Immediate Word Recall Task: “Please repeat all the words you can remember.”\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003ePerformance-based cognitive measure\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eb. Abstract Reasoning\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e• Number Pattern Task: “What number is missing: 2, 4, 6, __, 10?”\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003ePerformance-based cognitive measure\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003e3. Health Risk-Taking Behaviour (2014)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ea. Smoking\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e• “Have you ever chewed tobacco, smoked a pipe, smoked self-rolled cigarettes, or smoked cigarettes/cigars?”\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eDichotomous indicator\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eb. Unhealthy Eating\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e• Fast food, sweet snacks, soft drinks, fried food (in past 7 days)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eItems summed for total unhealthy eating score\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ec. Physical Inactivity\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e• “Did you do any vigorous activity for ≥ 10 minutes?” (past 7 days) \u003cem\u003e(Reverse-coded)\u003c/em\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eCaptures lack of activity\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003e4. Behavioural Change (2007–2014)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003ea. Smoking Change\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e• Still not smoking, Started smoking, Quit smoking, Still smoking\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eCategorical transition variable\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003eb. Physical Activity Change\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e• Remained active, Became inactive, Became active, Remained inactive\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003eCategorical transition variable\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e\u003cp\u003e\u003cb\u003eMissing Data and Imputation Strategy\u003c/b\u003e\u003c/p\u003e\u003cp\u003eGiven the longitudinal design and sample attrition between IFLS waves, missing data presented a key analytical challenge. To preserve sample size and reduce bias, we employed Multiple Imputation by Chained Equations (MICE) using the Random Forest (RF) method in R. This approach is especially suitable for datasets involving complex, non-linear relationships between variables, as in this study’s examination of social structure, cognition, and health behaviours (Doove et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2014\u003c/span\u003e; Jia \u0026amp; Wu, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Stekhoven \u0026amp; Bühlmann, \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Twenty-four of 36 key variables had missing values, and each was imputed based on all other available variables using a fully specified predictor matrix. Unique identifiers (e.g. PIDLINK) and fully observed fields were excluded. The RF method invoked via method = \"rf\" in the mice() function uses a flexible, non-parametric algorithm capable of handling categorical, ordinal, and continuous data. This enabled us to preserve variable-specific distributions and inter-variable relationships without relying on strong parametric assumptions (Bouhlila \u0026amp; Sellaouti, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). We generated five imputed datasets, each with 50 iterations per chain to ensure convergence, guided by best practices balancing computational efficiency and stability. A random seed (123) was set for reproducibility. Diagnostics confirmed convergence and consistency across imputations.\u003c/p\u003e\u003cp\u003eTo assess imputation quality, we compared key distributions pre- and post-imputation. Differences were minimal, suggesting that the imputation process preserved the plausibility and structure of the original data. All distributions remained within plausible bounds, and no implausible outliers or significant distortions were introduced. As a result, the final imputed dataset contained 11,539 complete cases across all 36 variables (0% missingness). Crucially, imputation enabled downstream structural equation modelling (SEM) that would otherwise be infeasible due to missingness. For instance, CFA and SEM could not produce valid estimates using the original dataset, whereas the imputed datasets yielded models with good fit indices (e.g., CFI and TLI), allowing robust estimation of the direct and indirect effects explored in this study. By applying MICE with RF, this study maintained analytical power, addressed potential biases from non-random missingness, and ensured that subsequent modelling of adolescent cognitive development and health behaviour trajectories reflected the full sample’s variability and complexity.\u003c/p\u003e\u003ch2\u003eStatistical Analysis\u003c/h2\u003e\u003cp\u003eThis study applied Structural Equation Modelling (SEM) to examine the role of cognitive status in shaping health risk behaviours among Indonesian adolescents and young adults. Three theory-driven SEM models were estimated:\u003c/p\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eModel 1: Cognitive status as a direct predictor of health risk-taking behaviour.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eModel 2: Cognitive status as a mediator between early-life social structure and health risk behaviour.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eModel 3: Cognitive status as a mediator between early-life social structure and behavioural change from adolescence to young adulthood.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003cp\u003eAll latent constructs were developed based on existing theory and prior empirical work. Given the established nature of the constructs and the emphasis of this study on examining structural pathways, we focused on model-level fit and structural pathways rather than reporting standalone Confirmatory Factor Analysis (CFA) results, which aligns with practices in theory-driven SEM applications in public health. All SEM analyses were conducted using the \u003cem\u003elavaan\u003c/em\u003e package in R. Models controlled for age and gender, and robust standard errors were applied to account for any non-normality. Model fit was evaluated using standard SEM indices, including the Comparative Fit Index (CFI), Tucker-Lewis Index (TLI), Root Mean Square Error of Approximation (RMSEA), and Standardised Root Mean Square Residual (SRMR).\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003e1. Descriptive Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSample Demographics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBefore examining the structural model, we first describe the socio-demographic characteristics and distribution of key variables to contextualise the analysis. The study sample consisted of 11,539 Indonesian adolescents and young adults drawn from the Indonesia Family Life Survey (IFLS). At the time of outcome measurement in 2014, participants were aged 22\u0026ndash;30 (mean\u0026thinsp;=\u0026thinsp;26.1, SD\u0026thinsp;=\u0026thinsp;2.4), having been 15\u0026ndash;23 years old in 2007 (mean\u0026thinsp;=\u0026thinsp;19.1, SD\u0026thinsp;=\u0026thinsp;2.4) (Table \u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). The gender distribution was 52.4% female and 47.6% male. The majority of participants (56.7%) resided in urban areas. Ethnically, 39.4% identified as Javanese, 11.9% as Sundanese, 4.9% as Batak, and the remainder from other groups. A large majority (89.6%) were Muslim, consistent with national demographics. Socio-economic indicators showed a wide range, with significant disparities between low- and high-income households. The average annual household income in 2007 was IDR 31.2 million, ranging from low-income to over IDR 2 billion. Mean land asset value was IDR 75.1 million, and vehicle assets averaged IDR 11 million. Approximately 42.4% of respondents had completed senior high school, while 10% attained higher education. Parental education levels were generally lower, with only 20.5% having completed education beyond senior high. The average distance to school was 13.25 km (SD\u0026thinsp;=\u0026thinsp;45.6), highlighting infrastructural barriers in rural areas. Regarding family structure, 93% of respondents lived with both parents at age 12, and 91.4% reported parental co-residence. Parental communication, rated on a 6\u0026ndash;30 scale, had a mean score of 14 (SD\u0026thinsp;=\u0026thinsp;4.4). Youth community engagement was uneven; only 26.6% reported active participation in community activities such as Karang Taruna.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eDemographic characteristics\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eNo.\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVariable\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eWeighted Mean (SD) / Weighted % (95% CI)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eDemographics\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGender (% female)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6041 (52.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAge (2007)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e19.1 (2.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAge (2014)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e26.1 (2.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eUrban residence\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6548 (56.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eReligion (% Muslim)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10,338 (89.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eEthnicity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ea. Javanese\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4543 (39.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eb. Sundanese\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1370 (11.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ec. Batak\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e569 (4.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ed. Others\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3959 (34.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eSocioeconomic Status (SES)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTotal Household Income (IDR)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e31.3\u0026nbsp;million (65\u0026nbsp;million)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLand Asset Value (IDR)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e75.1\u0026nbsp;million (152\u0026nbsp;million)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVehicle Asset Value (IDR)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e11\u0026nbsp;million (32\u0026nbsp;million)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eParental Education Level\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eUnschooled\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3 (0.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e11\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eKindergarten\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e723 (6.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e12\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eElementary School\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2065 (17.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eJunior High School\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5602 (48.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e14\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSenior High School\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e675 (5.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e15\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eD1/D2/D3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2360 (20.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBachelor\u0026rsquo;s Degree\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e106 (0.9%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMaster\u0026rsquo;s/PhD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5 (0.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eParticipant\u0026rsquo;s Education Level\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eUnschooled\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e118 (1.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eKindergarten\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1 (0.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eElementary School\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2041 (17.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eJunior High School\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2830 (24.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSenior High School\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4887 (42.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eD1/D2/D3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e478 (4.1%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eBachelor\u0026rsquo;s Degree\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1151 (10.0%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e25\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMaster\u0026rsquo;s/PhD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e33 (0.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e26\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eDistance to School (KM)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13.25 (45.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eFamily Structure (at age 12)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eParents Married (Yes)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10,826 (93.8%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLiving with Both Parents\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10,548 (91.4%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e29\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFrequency of Communication with Parents (1\u0026ndash;5 scale)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4.4 (1.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eYouth Community Participation\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e30\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eParticipation in Youth Activities (Yes)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3071 (26.6%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eAccess to Health Services\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e31\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eHealth Service Access Score\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.6 (1.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cstrong\u003eCognitive Status\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCognitive status was assessed in 2014 using two tasks: a memory recall test (scores 1\u0026ndash;10) measuring episodic memory, and a pattern reasoning test (scores 0\u0026ndash;7) assessing fluid intelligence. As shown in Table \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e, males scored slightly higher on memory (mean\u0026thinsp;=\u0026thinsp;5.76) than females (5.56), while females had marginally higher pattern reasoning scores (2.99 vs. 2.98). Memory scores declined modestly with age, from 5.68 in ages 15\u0026ndash;16 to 5.54 in ages 23\u0026ndash;24, whereas pattern reasoning remained relatively stable, peaking at 3.09 in the 19\u0026ndash;20 age group. These patterns align with established cognitive development trajectories and support the use of these measures in further modelling.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab3\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eThe distribution of scores for each cognitive measure, age and gender.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eGroup\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eN\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMean Memory Score\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eMean Pattern Score\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eGender\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e6041\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.76\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2.98\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5498\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2.99\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eAge Group\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e15\u0026ndash;16\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e790\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.68\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2.84\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e16\u0026ndash;17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1289\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.69\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.02\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e17\u0026ndash;18\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1348\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.00\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e18\u0026ndash;19\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1477\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.71\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.03\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e19\u0026ndash;20\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1324\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.74\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.09\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e20\u0026ndash;21\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1379\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.70\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.01\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e21\u0026ndash;22\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1395\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.63\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2.91\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e22\u0026ndash;23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1491\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2.95\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e23\u0026ndash;24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1046\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.54\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2.99\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cstrong\u003eHealth Risk-Behaviours\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHealth risk behaviours showed concerning patterns. In 2007, 22.3% of respondents smoked; by 2014, this rose to 36.5%, reflecting increased uptake during the transition to adulthood. Only 22.2% reported engaging in vigorous physical activity in 2014, slightly up from 19.5% in 2007. Unhealthy eating, measured on a 0\u0026ndash;4 scale, had a mean score of 2.2 (SD\u0026thinsp;=\u0026thinsp;0.8), indicating moderate consumption of unhealthy foods such as sweets, fast food, and soft drinks (Table \u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab4\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eHealth Risk-Taking Behaviour\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVariable\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eWeighted Mean (SD) / Weighted Percentage (95% CI), N\u0026thinsp;=\u0026thinsp;11,539\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSmoking 2014 (yes)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e4208 (36.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSmoking 2007 (yes)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2573 (22.3%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePhysical activity 2014 (yes)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2563 (22.2%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePhysical activity 2007 (yes)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2249 (19.5%)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eUnhealthy eating behaviour (scale 1\u0026ndash;4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2.2 (0.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eThe descriptive analysis reveals important behavioural patterns during the transition from adolescence to young adulthood. For smoking, the majority of respondents remained non-smokers across both waves (62.9%), which is encouraging. However, a substantial proportion either continued smoking (21.7%) or initiated smoking during the period (14.8%) (Table \u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e). Alarmingly, very few adolescents reported quitting smoking (0.6%), highlighting persistently low cessation rates. This finding is particularly concerning, as it underscores how difficult it is to quit once smoking behaviour has been established.\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab5\" border=\"1\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 5\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eFrequency of behaviour changes\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eNo.\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eStatus Category\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSmoking Status\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eFrequency (Smoking)\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ePhysical Activity Status\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eFrequency (Physical Activity)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eStill not engaging\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0\u0026thinsp;=\u0026thinsp;Still Not Smoking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7,258 (62.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0\u0026thinsp;=\u0026thinsp;Still Active\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e7,615 (66)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePositive change\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1\u0026thinsp;=\u0026thinsp;Quit Smoking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e73 (0.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1\u0026thinsp;=\u0026thinsp;Started Active\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1,675 (14.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNegative change\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2\u0026thinsp;=\u0026thinsp;Started Smoking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1,708 (14.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2\u0026thinsp;=\u0026thinsp;Quit Active\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1,361 (11.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eStill engaging\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3\u0026thinsp;=\u0026thinsp;Still Smoking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e2,500 (21.7%)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3\u0026thinsp;=\u0026thinsp;Still Not Active\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e888 (7.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eTotal\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e11,539\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026mdash;\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003e11,539\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e2. Structural Models\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel 1: Cognitive status and health Risk Behaviour\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel 2: Cognitive Status as a Mediator between Social Structure and Health Risk Behaviour\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel 3: Cognitive Status as a Mediator between Social Structure and Behavioural Change\u003c/strong\u003e\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis study examined the role of cognitive status as both a direct predictor and a mediator of health risk-taking behaviour among adolescents and young adults in Indonesia using SEM across three models. The findings highlight a developmental pathway where cognitive function partially explains how early social inequalities lead to healthier behavioural outcomes, offering new insight into the mechanisms of adolescent behavioural development in a developing country context marked by health inequalities and a rapid second demographic transition characterised by falling fertility and an increasing aging population. The descriptive data reveal considerable heterogeneity in the sample\u0026rsquo;s socioeconomic and demographic background. The average annual household income in 2007 was IDR 31.2\u0026nbsp;million, ranging widely from low-income households to over IDR 2\u0026nbsp;billion, with substantial variability in land and vehicle assets. Educational attainment showed that 42.4% of respondents completed senior high school and 10% attained higher education, whereas parental education levels were generally lower, indicating intergenerational disparities. The average distance to school was notably high at 13.25 km (SD\u0026thinsp;=\u0026thinsp;45.6), reflecting infrastructural and geographic barriers prevalent in Indonesia\u0026rsquo;s rural and remote areas (World Bank, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Family structures were predominantly stable, with over 90% living with both parents at age 12 and moderate parental communication levels. However, youth community engagement was limited, with only 26.6% participating in activities such as Karang Taruna. This socio-structural context frames the latent social structure construct and likely shapes cognitive development and health behaviour patterns, as supported by previous studies emphasizing the role of structural conditions in adolescent development (Bradley \u0026amp; Corwyn, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Evans \u0026amp; Kim, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eFirst, the analysis confirmed that cognitive status measured through memory recall and abstract reasoning was a statistically significant protective factor against risk behaviours such as smoking, unhealthy eating, and physical inactivity. These findings are in line with prior research in high-income contexts, which suggests that cognitive skills underpin health literacy, future planning, and self-regulation, all of which are essential to avoid risky health behaviours (Batty et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Hackman et al., \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Reyna \u0026amp; Farley, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). Similarly, studies from other developing settings also emphasize how self-regulatory strategies often shaped by social support systems can mediate risk behaviours in youth. For example, Gaspar de Matos et al., (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) found that children and adolescents with stronger self-regulation and social support had greater healthy eating awareness, highlighting the importance of individual cognitive capacities reinforced by environmental conditions.\u003c/p\u003e\u003cp\u003eSecond, cognitive status emerged as a meaningful mediator in the relationship between early-life social structure and later health risk behaviours. Adolescents embedded in structurally advantaged environments characterised by higher parental education, household income, family stability, and community participation developed stronger cognitive capacities by young adulthood. These capacities, in turn, reduced their engagement in health risk behaviours. Although the mediation was partial, the pathway is consistent with life-course and developmental models positing cognition as a key mechanism linking structural conditions to behavioural health outcomes (Ferrer \u0026amp; McArdle, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2010\u003c/span\u003e; Halfon \u0026amp; Hochstein, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Shonkoff et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). However, it is important to note that the variance explained in cognitive status (R\u0026sup2; = 0.007) was relatively low, indicating that factors beyond those included in the structural model likely shape adolescent cognitive development. Early-life nutrition, psychosocial stimulation, and exposure to environmental stressors may play a crucial role in shaping cognitive outcomes (Grantham-McGregor et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2007\u003c/span\u003e; Noble et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2005\u003c/span\u003e). This reinforces the importance of holistic interventions that integrate health, nutrition, and educational support during childhood and adolescence. This perspective aligns with qualitative research from the Democratic Republic of Congo, which underscores how poverty, peer influence, and parental behaviour shape youth involvement in risky behaviours such as alcohol use and violence often through reduced self-regulatory and social protective factors (Kohli et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThird, the comparison of the three structural models provides important insights into how cognitive status operates in different behavioural contexts. Model 1 demonstrated that cognition had a direct protective effect on static health risk behaviours in young adulthood. In Model 2, cognition functioned as a partial mediator, linking early social advantage to reduced risk behaviours in 2014. However, in Model 3 focused on behavioural change over time cognition\u0026rsquo;s effect was mainly indirect, while early social structure had a stronger total effect on behavioural transitions. This highlights a critical distinction: cognitive skills may play a more direct role in supporting immediate health decisions (e.g., avoiding smoking), but structural factors are more powerful in influencing long-term behaviour patterns (e.g., quitting smoking or maintaining physical activity).\u003c/p\u003e\u003cp\u003eThe contrast between Model 2 and Model 3 is especially relevant. In Model 2, cognitive function was significantly associated with contemporaneous health risk behaviours, even after accounting for social structure, suggesting its role in enabling informed decision-making and behavioural restraint. In contrast, Model 3 showed that cognitive function played an indirect role, mediating the influence of early social advantage on changes in health behaviours across time. The stronger total effect of social structure in Model 3 suggests that long-term behavioural shifts are more deeply shaped by socioeconomic conditions than by cognitive capacity alone. This distinction adds depth to our understanding of how and when cognitive status matters most, highlighting that cognitive skills are critical for sustaining good behaviour but less influential when individuals are undergoing major behavioural transitions influenced by external contexts. These findings align with Bourdieu\u0026rsquo;s theory of habitus, where early-life structural positions shape durable dispositions and behavioural routines (Bourdieu, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e1977\u003c/span\u003e). This may help explain why structural conditions had a stronger effect on behaviour change: once established, behaviours become embedded in social environments that are difficult to shift through individual cognitive effort alone. The results also provide empirical support for the cumulative disadvantage framework, which posits that early structural inequalities compound over time to produce long-term disparities in health outcomes (Dannefer, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). Gender-specific pathways also emerged in the analysis, with males significantly more likely to engage in risky health behaviours. This reflects gendered norms in Indonesia that normalise smoking among boys while discouraging it among girls (Fithria et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Kodriati et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Such patterns suggest the need for gender-sensitive health interventions that address social norms alongside individual cognition.\u003c/p\u003e\u003cp\u003eThis layered comparison across models strengthens the originality of the study and demonstrates the utility of using SEM to parse these complex developmental and structural interactions in youth health research. Such modelling approaches remain rare in studies of adolescents in developing countries like Indonesia, where demographic shifts and structural inequalities co-exist. This study contributes to a more nuanced life-course understanding of adolescent health. From a policy perspective, the findings reinforce the value of integrated interventions. Programmes that invest in early-life structural conditions such as improving access to education, reducing poverty, promoting family cohesion, and enhancing community engagement can yield dual dividends: they promote cognitive development and reduce long-term health risks (Bradley \u0026amp; Corwyn, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2002\u003c/span\u003e; Marmot et al., \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2008\u003c/span\u003e). While targeted cognitive development initiatives such as programs aimed at improving memory, abstract reasoning, and decision-making may offer valuable support for youth health behaviours, they should not be pursued in isolation. Without addressing broader structural inequalities, such interventions risk disproportionately benefiting more advantaged adolescents, potentially exacerbating existing disparities. For instance, schools in low-resource settings may lack the infrastructure to implement cognitive enhancement programs, and young people in remote or underserved areas may have limited access to such opportunities. Research shows that socioeconomic disadvantage is closely associated with lower performance in memory and reasoning tasks, which may reduce the effectiveness of cognitive interventions unless underlying conditions are also addressed (Noble et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; von Stumm, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). Therefore, cognitive programming must be implemented alongside structural reforms such as investments in equitable education, nutrition, and health services to ensure access, effectiveness, and cultural relevance across diverse settings (Blair \u0026amp; Raver, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; OECD, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn Indonesia, these findings hold particular relevance given the country's ongoing demographic bonus, a growing young population and persistent socioeconomic disparities in health and education access. Scalable strategies could include embedding cognitive training in school curricula, supporting youth engagement platforms like Karang Taruna, and expanding adolescent access to preventive health and educational services. However, these efforts must be tailored to the local sociocultural context. For instance, given gender norms that stigmatise smoking among girls but normalise it for boys, behaviour change strategies should account for such social pressures. Furthermore, rural-urban disparities in health access and education may limit the reach of interventions unless accompanied by broader structural reforms. Addressing these inequalities is essential for achieving equitable outcomes during this critical demographic transition. These results offer timely and relevant evidence for adolescent health policy in Indonesia and similar developing countries. They affirm the role of cognitive development as a cross-cutting determinant of health and illustrate how early social context sets the foundation for youth health trajectories. The integration of cognitive and structural interventions could represent a cost-effective, life-course approach to reducing behavioural health risks among the next generation.\u003c/p\u003e\u003cp\u003eNevertheless, several limitations must be noted. First, although the study employs a longitudinal design, unmeasured confounding factors such as personality traits or mental health status could bias the relationships observed. Second, the cognitive measures used for memory and pattern recognition do not capture the full range of executive functions that may influence health behaviour. Third, the indicators of health risk behaviour were limited to smoking, diet, and physical activity, omitting other relevant behaviours such as alcohol use or sexual risk. Finally, the behavioural change variable was constructed from self-reported ordinal data, which may be subject to recall bias or social desirability effects. Future research should consider incorporating a broader array of cognitive, behavioural, and psychosocial indicators to better capture the complexity of adolescent development. It would also be valuable to examine how specific community-level factors, such as peer norms, school quality, or local health services, moderate the cognitive-behaviour relationship. Mixed-methods approaches could enrich understanding of how adolescents interpret and act on health risks in structurally diverse settings. Ultimately, these refinements could enhance the effectiveness and equity of adolescent health interventions in Indonesia and other developing settings.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study contributes new evidence to understanding the role of cognitive development in adolescent health behaviour within a developing country context. \u0026nbsp;Using three SEM models, it establishes that cognitive status both directly and indirectly shapes risk-taking behaviour, and partially explains the impact of early social advantage on later outcomes. Importantly, the study highlights that cognition has a more pronounced effect on contemporaneous health decisions, while long-term behavioural changes are more strongly influenced by structural conditions. These findings support the need for a dual-focus strategy that combines cognitive development efforts with broader structural reforms. In the Indonesian context marked by demographic opportunity but persistent inequalities such an integrated approach could help support healthier youth transitions. However, to ensure these interventions do not inadvertently widen disparities, they must be equitably designed, resourced, and sensitive to the sociocultural and geographic diversity of the population.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthical approval statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study used de-identified secondary data from the Indonesia Family Life Survey (IFLS), which received ethical approval from the Institutional Review Boards at RAND and the University of Gadjah Mada. Additional ethical clearance for this analysis was obtained through the University of Sheffield\u0026rsquo;s ethics application system (Reference Number: 057057). All procedures were conducted in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo funding was received for conducting this study\u003cstrong\u003e.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research was supported by The Indonesian Education Scholarship Program (Beasiswa Pendidikan Indonesia - BPI), funded by the Centre for Higher Education Funding and Assessment, Ministry of Higher Education, Science, and Technology of the Republic of Indonesia. The author gratefully acknowledges this support.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDisclosure statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo potential conflict interest was reported by the authors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used in this study are publicly available in the Indonesia Family Life Survey (IFLS) at https://www.rand.org/well-being/social-and-behavioral-policy/data/FLS/IFLS/ifls5.html. The dataset analysed during the current study is available from correspondence author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical trial registration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eClinical trial registration: Not applicable\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eArnett, J. J. (2000). Emerging adulthood: A theory of development from the late teens through the twenties. \u003cem\u003eAmerican Psychologist\u003c/em\u003e, \u003cem\u003e55\u003c/em\u003e(5), 469\u0026ndash;480. https://doi.org/10.1037/0003-066X.55.5.469\u003c/li\u003e\n \u003cli\u003eBatty, G. D., Deary, I. J., \u0026amp; Gottfredson, L. S. (2007). Premorbid (early life) IQ and later mortality risk: Systematic review. \u003cem\u003eAnnals of Epidemiology\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(4), 278\u0026ndash;288. https://doi.org/10.1016/j.annepidem.2006.07.010\u003c/li\u003e\n \u003cli\u003eBlair, C., \u0026amp; Raver, C. C. (2016). Poverty, Stress, and Brain Development: New Directions for Prevention and Intervention. \u003cem\u003eAcademic Pediatrics\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(3 Suppl), S30-36. https://doi.org/10.1016/j.acap.2016.01.010\u003c/li\u003e\n \u003cli\u003eBouhlila, D. S., \u0026amp; Sellaouti, F. (2013). Multiple imputation using chained equations for missing data in TIMSS: A case study. \u003cem\u003eLarge-Scale Assessments in Education\u003c/em\u003e, \u003cem\u003e1\u003c/em\u003e(1), 4. https://doi.org/10.1186/2196-0739-1-4\u003c/li\u003e\n \u003cli\u003eBourdieu, P. (with Internet Archive). (1977). \u003cem\u003eOutline of a theory of practice\u003c/em\u003e. Cambridge ; New York : Cambridge University Press. http://archive.org/details/outlineoftheoryo0000bour\u003c/li\u003e\n \u003cli\u003eBradley, R. H., \u0026amp; Corwyn, R. F. (2002). Socioeconomic status and child development. \u003cem\u003eAnnual Review of Psychology\u003c/em\u003e, \u003cem\u003e53\u003c/em\u003e(1), 371\u0026ndash;399. https://doi.org/10.1146/annurev.psych.53.100901.135233\u003c/li\u003e\n \u003cli\u003eBrieant, A., Peviani, K. M., Lee, J. E., King-Casas, B., \u0026amp; Kim-Spoon, J. (2021). Socioeconomic Risk for Adolescent Cognitive Control and Emerging Risk-Taking Behaviors. \u003cem\u003eJournal of Research on Adolescence\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e(1), 71\u0026ndash;84. https://doi.org/10.1111/jora.12583\u003c/li\u003e\n \u003cli\u003eClark, C., Rodgers, B., Caldwell, T., Power, C., \u0026amp; Stansfeld, S. (2007). Childhood and adulthood psychological ill health as predictors of midlife affective and anxiety disorders: The 1958 British Birth Cohort. \u003cem\u003eArchives of General Psychiatry\u003c/em\u003e, \u003cem\u003e64\u003c/em\u003e(6), 668\u0026ndash;678. https://doi.org/10.1001/archpsyc.64.6.668\u003c/li\u003e\n \u003cli\u003eCosta, P. T., McCrae, R. R., \u0026amp; Dye, D. A. (1991). Facet Scales for Agreeableness and Conscientiousness: A Revision of the NEO Personality Inventory. \u003cem\u003ePersonality and Individual Differences\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(9), 887\u0026ndash;898. https://doi.org/10.1016/0191-8869(91)90177-D\u003c/li\u003e\n \u003cli\u003eDaly, A. N., O\u0026rsquo;Sullivan, E. J., \u0026amp; Kearney, J. M. (2022). Considerations for health and food choice in adolescents. \u003cem\u003eProceedings of the Nutrition Society\u003c/em\u003e, \u003cem\u003e81\u003c/em\u003e(1), 75\u0026ndash;86. https://doi.org/10.1017/S0029665121003827\u003c/li\u003e\n \u003cli\u003eDannefer, D. (2003). Cumulative advantage/disadvantage and the life course: Cross-fertilizing age and social science theory. \u003cem\u003eThe Journals of Gerontology. Series B, Psychological Sciences and Social Sciences\u003c/em\u003e, \u003cem\u003e58\u003c/em\u003e(6), S327-337. https://doi.org/10.1093/geronb/58.6.s327\u003c/li\u003e\n \u003cli\u003eDoove, L. L., Van Buuren, S., \u0026amp; Dusseldorp, E. (2014). Recursive partitioning for missing data imputation in the presence of interaction effects. \u003cem\u003eComputational Statistics \u0026amp; Data Analysis\u003c/em\u003e, \u003cem\u003e72\u003c/em\u003e, 92\u0026ndash;104. https://doi.org/10.1016/j.csda.2013.10.025\u003c/li\u003e\n \u003cli\u003eEvans, G. W., \u0026amp; Kim, P. (2013). Childhood poverty, chronic stress, self‐regulation, and coping. \u003cem\u003eChild Development Perspectives\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(1), 43\u0026ndash;48. https://doi.org/10.1111/cdep.12013\u003c/li\u003e\n \u003cli\u003eFerrer, E., \u0026amp; McArdle, J. J. (2010). Longitudinal modeling of developmental changes in psychological research. \u003cem\u003eCurrent Directions in Psychological Science\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(3), 149\u0026ndash;154. https://doi.org/10.1177/0963721410370300\u003c/li\u003e\n \u003cli\u003eFithria, F., Adlim, M., Jannah, S. R., \u0026amp; Tahlil, T. (2021). Indonesian adolescents\u0026rsquo; perspectives on smoking habits: A qualitative study. \u003cem\u003eBMC Public Health\u003c/em\u003e, \u003cem\u003e21\u003c/em\u003e(1), 82. https://doi.org/10.1186/s12889-020-10090-z\u003c/li\u003e\n \u003cli\u003eGaspar de Matos, M., Palmeira, A. L., Gaspar, T., De Wit, J. B. F., \u0026amp; Luszczynska, A. (2016). Social support influences on eating awareness in children and adolescents: The mediating effect of self-regulatory strategies. \u003cem\u003eGlobal Public Health\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(4), 437\u0026ndash;448. https://doi.org/10.1080/17441692.2015.1094106\u003c/li\u003e\n \u003cli\u003eGrantham-McGregor, S., Cheung, Y. B., Cueto, S., Glewwe, P., Richter, L., Strupp, B., \u0026amp; International Child Development Steering Group. (2007). Developmental potential in the first 5 years for children in developing countries. \u003cem\u003eLancet (London, England)\u003c/em\u003e, \u003cem\u003e369\u003c/em\u003e(9555), 60\u0026ndash;70. https://doi.org/10.1016/S0140-6736(07)60032-4\u003c/li\u003e\n \u003cli\u003eHackman, D. A., Farah, M. J., \u0026amp; Meaney, M. J. (2010). Socioeconomic status and the brain: Mechanistic insights from human and animal research. \u003cem\u003eNature Reviews Neuroscience\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e(9), 651\u0026ndash;659.\u003c/li\u003e\n \u003cli\u003eHalfon, N., \u0026amp; Hochstein, M. (2002). Life course health development: An integrated framework for developing health, policy, and research. \u003cem\u003eThe Milbank Quarterly\u003c/em\u003e, \u003cem\u003e80\u003c/em\u003e(3), 433\u0026ndash;479, iii. https://doi.org/10.1111/1468-0009.00019\u003c/li\u003e\n \u003cli\u003eJia, F., \u0026amp; Wu, W. (2023). A comparison of multiple imputation strategies to deal with missing nonnormal data in structural equation modeling. \u003cem\u003eBehavior Research Methods\u003c/em\u003e, \u003cem\u003e55\u003c/em\u003e(6), 3100\u0026ndash;3119. https://doi.org/10.3758/s13428-022-01936-y\u003c/li\u003e\n \u003cli\u003eJunger, M., \u0026amp; van Kampen, M. (2010). Cognitive ability and self-control in relation to dietary habits, physical activity and bodyweight in adolescents. \u003cem\u003eThe International Journal of Behavioral Nutrition and Physical Activity\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e, 22. https://doi.org/10.1186/1479-5868-7-22\u003c/li\u003e\n \u003cli\u003eKodriati, N., Hayati, E. N., Santosa, A., \u0026amp; Pursell, L. (2020). Perceived social benefits versus perceived harms of smoking among Indonesian boys aged 12\u0026ndash;16 years: A secondary analysis of Global Youth Tobacco Survey 2014. \u003cem\u003eTobacco Prevention \u0026amp; Cessation\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e, 8. https://doi.org/10.18332/tpc/115034\u003c/li\u003e\n \u003cli\u003eKohli, A., Remy, M. M., Binkurhorhwa, A. K., Mitima, C. M., Mirindi, A. B., Mwinja, N. B., Banyewesize, J. H., Ntakwinja, G. M., Perrin, N. A., \u0026amp; Glass, N. (2018). Preventing risky behaviours among young adolescents in eastern Democratic Republic of Congo: A qualitative study. \u003cem\u003eGlobal Public Health\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(9), 1241\u0026ndash;1253. https://doi.org/10.1080/17441692.2017.1317009\u003c/li\u003e\n \u003cli\u003eMarmot, M., Friel, S., Bell, R., Houweling, T. A. J., Taylor, S., \u0026amp; Commission on Social Determinants of Health. (2008). Closing the gap in a generation: Health equity through action on the social determinants of health. \u003cem\u003eLancet (London, England)\u003c/em\u003e, \u003cem\u003e372\u003c/em\u003e(9650), 1661\u0026ndash;1669. https://doi.org/10.1016/S0140-6736(08)61690-6\u003c/li\u003e\n \u003cli\u003eNg, M., Freeman, M. K., Fleming, T. D., Robinson, M., Dwyer-Lindgren, L., Thomson, B., Wollum, A., Sanman, E., Wulf, S., Lopez, A. D., Murray, C. J. L., \u0026amp; Gakidou, E. (2014). Smoking prevalence and cigarette consumption in 187 countries, 1980-2012. \u003cem\u003eJAMA\u003c/em\u003e, \u003cem\u003e311\u003c/em\u003e(2), 183\u0026ndash;192. https://doi.org/10.1001/jama.2013.284692\u003c/li\u003e\n \u003cli\u003eNoble, K. G., Norman, M. F., \u0026amp; Farah, M. J. (2005). Neurocognitive correlates of socioeconomic status in kindergarten children. \u003cem\u003eDevelopmental Science\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(1), 74\u0026ndash;87. https://doi.org/10.1111/j.1467-7687.2005.00394.x\u003c/li\u003e\n \u003cli\u003eOECD. (2025). \u003cem\u003eReducing Inequalities by Investing in Early Childhood Education and Care: Project methodology for translating research into early childhood education and care policies\u003c/em\u003e. OECD. https://www.oecd.org/en/publications/reducing-inequalities-by-investing-in-early-childhood-education-and-care_b78f8b25-en/full-report/project-methodology-for-translating-research-into-early-childhood-education-and-care-policies_93f1eb79.html\u003c/li\u003e\n \u003cli\u003ePatton, G. C., Sawyer, S. M., Santelli, J. S., Ross, D. A., Afifi, R., Allen, N. B., Arora, M., Azzopardi, P., Baldwin, W., Bonell, C., Kakuma, R., Kennedy, E., Mahon, J., McGovern, T., Mokdad, A. H., Patel, V., Petroni, S., Reavley, N., Taiwo, K., \u0026hellip; Viner, R. M. (2016). Our future: A Lancet commission on adolescent health and wellbeing. \u003cem\u003eThe Lancet\u003c/em\u003e, \u003cem\u003e387\u003c/em\u003e(10036), 2423\u0026ndash;2478. https://doi.org/10.1016/S0140-6736(16)00579-1\u003c/li\u003e\n \u003cli\u003eReyna, V. F., \u0026amp; Farley, F. (2006). Risk and rationality in adolescent decision making: Implications for theory, practice, and public policy. \u003cem\u003ePsychological Science in the Public Interest\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e(1), 1\u0026ndash;44.\u003c/li\u003e\n \u003cli\u003eSeptiono, W., Kuipers, M. A. G., Ng, N., \u0026amp; Kunst, A. E. (2022). Self-reported exposure of Indonesian adolescents to online and offline tobacco advertising, promotion and sponsorship (TAPS). \u003cem\u003eTobacco Control\u003c/em\u003e, \u003cem\u003e31\u003c/em\u003e(1), 98\u0026ndash;105. https://doi.org/10.1136/tobaccocontrol-2020-056080\u003c/li\u003e\n \u003cli\u003eShonkoff, J. P., Boyce, W. T., \u0026amp; McEwen, B. S. (2009). Neuroscience, molecular biology, and the childhood roots of health disparities: Building a new framework for health promotion and disease prevention. \u003cem\u003eJAMA\u003c/em\u003e, \u003cem\u003e301\u003c/em\u003e(21), 2252\u0026ndash;2259. https://doi.org/10.1001/jama.2009.754\u003c/li\u003e\n \u003cli\u003eSteinberg, L. (2008). A social neuroscience perspective on adolescent risk-taking. \u003cem\u003eDevelopmental Review\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e(1), 78\u0026ndash;106. https://doi.org/10.1016/j.dr.2007.08.002\u003c/li\u003e\n \u003cli\u003eStekhoven, D. J., \u0026amp; B\u0026uuml;hlmann, P. (2012). MissForest\u0026mdash;Nonparametric missing value imputation for mixed-type data. \u003cem\u003eBioinformatics\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e(1), 112\u0026ndash;118. https://doi.org/10.1093/bioinformatics/btr597\u003c/li\u003e\n \u003cli\u003eStrauss, J., Witoelar, F., \u0026amp; Sikoki, B. (2016). \u003cem\u003eThe Fifth Wave of the Indonesia Family Life Survey: Overview and Field Report: Volume 1\u003c/em\u003e. RAND Corporation. https://doi.org/10.7249/WR1143.1\u003c/li\u003e\n \u003cli\u003eUNFPA. (2025). \u003cem\u003eWorld Population Dashboard -Indonesia | United Nations Population Fund\u003c/em\u003e. https://www.unfpa.org/data/world-population/ID\u003c/li\u003e\n \u003cli\u003eViner, R. M., Ozer, E. M., Denny, S., Marmot, M., Resnick, M., Fatusi, A., \u0026amp; Currie, C. (2012). Adolescence and the social determinants of health. \u003cem\u003eLancet (London, England)\u003c/em\u003e, \u003cem\u003e379\u003c/em\u003e(9826), 1641\u0026ndash;1652. https://doi.org/10.1016/S0140-6736(12)60149-4\u003c/li\u003e\n \u003cli\u003evon Stumm, S. (2017). Socioeconomic status amplifies the achievement gap throughout compulsory education independent of intelligence. \u003cem\u003eIntelligence\u003c/em\u003e, \u003cem\u003e60\u003c/em\u003e, 57\u0026ndash;62. https://doi.org/10.1016/j.intell.2016.11.006\u003c/li\u003e\n \u003cli\u003eWhalley, L. J., \u0026amp; Deary, I. J. (2001). \u003cem\u003eLongitudinal cohort study of childhood IQ and survival up to age 76\u003c/em\u003e. https://doi.org/10.1136/bmj.322.7290.819\u003c/li\u003e\n \u003cli\u003eWorld Bank. (2017). \u003cem\u003eImproving Education Quality in Indonesia\u0026rsquo;s Poor Rural and Remote Areas\u003c/em\u003e [Text/HTML]. World Bank. https://www.worldbank.org/en/results/2017/12/22/improving-education-quality-in-indonesia-poor-rural-and-remote-areas\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-public-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pubh","sideBox":"Learn more about [BMC Public Health](http://bmcpublichealth.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/pubh/default.aspx","title":"BMC Public Health","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"cognitive, social structure, adolescence, health risk-behaviour, structural modelling, longitudinal study","lastPublishedDoi":"10.21203/rs.3.rs-7244173/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7244173/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground:\u003c/h2\u003e\u003cp\u003eCognitive development during adolescence plays a critical role in shaping decision-making and self-regulation. However, its role in influencing health risk behaviours particularly as a pathway linking early-life social conditions to later health outcomes remains underexplored in middle-income settings. This study investigates how cognitive status both directly and indirectly influences health risk behaviours among Indonesian youth, with a focus on the mediating role of cognition between early social structure and later health behaviours.\u003c/p\u003e\u003ch2\u003eMethods:\u003c/h2\u003e\u003cp\u003eUsing longitudinal data from the Indonesia Family Life Survey (2007\u0026ndash;2014), which tracked individuals aged 15\u0026ndash;30, we applied structural equation modelling (SEM) to test three hypotheses: (1) the direct effect of adolescent cognitive status on health risk behaviours in young adulthood; (2) its mediating role between early-life social structure and risk behaviour; and (3) its contribution to behavioural change over time in smoking and physical activity.\u003c/p\u003e\u003ch2\u003eResults:\u003c/h2\u003e\u003cp\u003eHigher cognitive status in adolescence was significantly associated with lower engagement in health risk behaviours such as smoking, unhealthy diet, and physical inactivity in young adulthood (standardized coefficient = -0.12, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Cognitive status partially mediated the relationship between early social advantage and health risk behaviours (indirect effect = -0.06, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001), while the direct effect of social structure remained significant (direct effect = -0.08, p\u0026thinsp;\u0026lt;\u0026thinsp;0.01). Although cognitive status did not independently predict behavioural change over time, its indirect influence through social structure remained substantial (indirect effect = -0.22, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001).Keywords: health risk behaviour, adolescents, youth Indonesia, cognitive status, social structure.\u003c/p\u003e\u003ch2\u003eConclusions:\u003c/h2\u003e\u003cp\u003eCognitive development plays a dual role as a product of social conditions and a protective factor against health risk behaviours. In rapidly developing countries like Indonesia, policies that enhance cognitive development alongside efforts to improve social-economic conditions may be particularly effective in supporting healthier behaviours during the transition from adolescence to young adulthood.\u003c/p\u003e","manuscriptTitle":"Understanding Youth Health Risk Behaviours in Indonesia through a Structural Model of Social and Cognitive Determinants: A Longitudinal Analysis","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-19 12:35:57","doi":"10.21203/rs.3.rs-7244173/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"editorInvitedReview","content":"","date":"2025-09-13T02:14:49+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-10T11:04:57+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"281679409226200547391293868614456801808","date":"2025-08-21T00:51:39+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"335796345216523169136186906566983635104","date":"2025-08-13T19:42:56+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-08-11T15:00:52+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-08-01T06:11:23+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-07-31T10:58:33+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-07-31T10:56:40+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Public Health","date":"2025-07-29T14:00:57+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-public-health","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"pubh","sideBox":"Learn more about [BMC Public Health](http://bmcpublichealth.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/pubh/default.aspx","title":"BMC Public Health","twitterHandle":"@BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ccf0d6fc-666c-4193-ad84-8d9b6f70e663","owner":[],"postedDate":"August 19th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2025-08-19T12:35:57+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-19 12:35:57","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7244173","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7244173","identity":"rs-7244173","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}