Latent Growth Curve Modeling Analyses Research Articles

A Longitudinal Study of the Development of Executive Function and Calibration Accuracy.

This longitudinal study examined the development of executive function and calibration accuracy in preadolescents. This study's sample consisted of 262 students (127 females) from grades 4 (n = 91), 5 (n = 89), and 6 (n = 82) who took measures of executive function and performance calibration in a sport task three times over 20 months. A latent growth-curve modeling analysis showed a significant relationship between the rates of change of executive function and calibration accuracy. The results also showed a dynamic interplay in the development of executive function and calibration accuracy. There were significant interindividual differences in the estimated population means both in executive function and calibration accuracy and in the rate of change of executive function, but not in the rate of change of calibration accuracy. The age of the participants had a positive effect only on the estimated population mean of executive function.

Practice-based evidence for spiritually integrated psychotherapies: Examining trajectories of psychological and spiritual distress.

The aims of this practice-based evidence study were to (a) examine clients' trajectories of psychological and spiritual distress over the course of spiritually integrated psychotherapies (SIPs) and (b) explore the role of varying types of spiritual interventions in these outcomes. In total, 164 practitioners of SIPs from 37 settings in a practice-research network administered the Clinically Adaptive Multidimensional Outcome Survey (Sanders et al., 2018) at each session with 1,227 clients and reported their use of theoretical orientations and spiritual interventions on an after-session summary checklist. Focusing on sessions over an initial 12-week period, latent growth curve modeling analyses revealed that clients, on average, experienced significant reduction of psychological distress during their engagement in SIPs with improvements occurring most sharply in the first month. Further, other findings revealed a salient reciprocal interplay with spiritual distress throughout treatment, such that clients who were struggling with their religious faith and/or spirituality were more psychologically distressed and displayed a more attenuated and gradual pattern of symptom reduction. In such cases, clinicians frequently utilized spiritual interventions involving basic skills (e.g., spiritual assessment), virtues (e.g., discuss self-control), and religious attachment (e.g., encourage acceptance of divine love) that were uniquely associated with clients' rate and duration of decline in psychological and spiritual distress. The present findings affirm the routine effectiveness of SIPs along with highlighting the potential value of certain spiritual interventions in supporting holistic recovery among clients who want clinicians to be culturally responsive to their spiritual and/or religious identities. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Flourish, fight or flight: health in self-employment over time—associations with individual and business resources

PurposeUsing COR theory to study developments of health and other key resources in self-employed workers in Sweden over 6 years, this study: (1) explored whether the heterogenous group of self-employed workers contained subgroups with different health trajectories, (2) investigated whether these were more typical for certain individuals (with respect to age, gender, sector, education, employment status), and (3) compared the different health trajectories regarding resource development in mental well-being, business resources, employment status, work ability.MethodThe study used data from the Swedish longitudinal occupational survey of health (SLOSH) and included participants working as self-employed or combiner (N = 2642).ResultFive trajectories were identified with latent class growth curve model analysis (LCGM). Two health trajectories with (1) very good, respective (2) good stable health (together comprising 78.5% of the participants), (3) one with moderate stable health (14.8%), (4) one with a U-shaped form (1.9%), and (5) one with low, slightly increasing health (4.7%). The first two trajectories flourish: they maintained or increased in all key resources and were more likely to remain self-employed. Trajectories three and five consist of those who fight to maintain or increase their resources. Workers in the U-shaped health trajectory show signs of fight and flight after loss in health and other key resources.ConclusionsStudying subgroups with different resource developments over time was suitable to understand heterogeneity in self-employed workers. It also helped to identify vulnerable groups that may benefit from interventions to preserve their resources.

Central systolic blood pressure as a risk for the accelerated progression of arterial stiffness

Abstract In the arterial tree, the systolic blood pressure (SBP) exhibits a pressure gradient from central to peripheral and is higher in the periphery because of pressure wave reflection. However, this gradient is attenuated with the elevation of central SBP (cSBP) in case with abnormal pressure wave reflection in arterial tree. It remains unclear whether increase of the cSBP might be an independent risk for the accelerated progression of arterial stiffness. Using the latent growth curve model (LGCM) analysis, this prospective observational study was conducted to examine whether elevation of the cSBP might be an independent risk for the accelerated progression of arterial stiffness in middle-aged Japanese men.In this 9-year prospective observational study, 3862 middle-aged Japanese men (42 ± 9 years old) without cerebrocardiovascular disease at the study baseline underwent annual measurements of the brachial-ankle pulse wave velocity (baPWV), and cSBP, estimated by radial pressure wave analysis as the second peak of radial pressure wave form (SBP2). During the follow-up period (6.4 ± 2.5), significant increases of baPWV and SBP2 were observed in all subjects. Analysis using the LGCM confirmed that the SBP2, a maker of cSBP (B=0.260, P < 0.001) was significant determinants of the slope of annual changes of baPWV during the study period. Our finding suggests that elevated cSBP is an independent risk for the accelerated progression of arterial stiffness in middle-aged Japanese men.

27 Risky Decision-Making Moderates the Association Between Motives for Cannabis Use and Cannabis Use Trajectories Among Adolescents

Objective:Prior literature has documented how motives for cannabis use predict frequency of use and cannabis use problems among adolescents. However, few studies have examined possible moderating variables that may influence the association between cannabis use motives and frequency of use. The current study examines how risky decision-making moderates this association to help better understand which individuals are at greater risk for cannabis use escalation. The current study will be the first to examine the interactive effects of motives for cannabis use (i.e., health or recreational reasons) and risky decision-making on cannabis use trajectories among a sample of adolescent cannabis users.Participants and Methods:Data from 194 adolescent cannabis users aged 14–17 at baseline were analyzed as part of a larger longitudinal study. Participants included those who self-reported use of cannabis within six months prior to the baseline assessment. The Marijuana Reasons for Use Questionnaire (MJRUQ) was used to assess motives for cannabis use from a list of 13 items. A confirmatory factor analysis identified “health” and “recreational” factors for motives for cannabis use. Lifetime frequency of cannabis use (number of days used) was assessed through the Drug Use History Questionnaire, while risky decision-making was assessed using the Game of Dice Task. We used latent growth curve modeling and linear regression analyses to examine the interactive effects of motives for cannabis use and risky decision-making on initial levels of lifetime cannabis use at baseline, and rate of cannabis use escalation over time.Results:No significant interactive effects were found for health motives for cannabis use; however, we found significant main effects of health motives on initial levels of lifetime cannabis use at baseline (b = 100.82, p < .01) and rate of cannabis use escalation (b = 24.79, p < .01). Those with a greater proclivity to use cannabis for health purposes showed higher initial levels of lifetime use at baseline and steeper increases in the rate of cannabis use escalation relative to those less likely to use for health purposes. Furthermore, we found a significant interactive effect of recreational motives for use and risky decision-making on the rate of cannabis use escalation (b = -2.53, p < .01). Follow-up analyses revealed that among those less likely to use cannabis for recreational purposes, higher risky decision-making was associated with a steeper increase in the rate of cannabis use escalation relative to those who exhibited lower risky decision-making.Conclusions:The current study replicated findings suggesting that cannabis use motives influence cannabis use trajectories. We found that using cannabis primarily for health reasons was associated with higher initial levels and steeper increases in use regardless of decision-making. Furthermore, we found that both motives for use and risky decision-making interacted to influence associations with cannabis use trajectories. Specifically, among individuals reporting less cannabis use for recreational reasons, those with relatively riskier decision-making showed steeper increases in the rate of cannabis use escalation. These findings inform prevention and intervention practices that focus on decision-making by tailoring approaches based on an individual’s primary motives for cannabis use.

Motives for Cannabis Use and Risky Decision Making Influence Cannabis Use Trajectories in Teens.

The current study will examine the interactive effects of motives for cannabis use (i.e., health or recreational) and risky decision making (DM) on cannabis use trajectories among adolescents. Data from 171 adolescents, aged 14-17 at the initial visit (baseline), were prospectively analyzed across five time points approximately six months apart. Latent growth curve modeling and linear regression analyses were used. We found a significant interactive effect of "recreational motives" and risky DM on the rate of cannabis use over time. Specifically, among those less likely to use cannabis for recreational purposes, riskier DM was associated with a faster increase in the rate of use over time relative to those with lower risky DM. Additionally, a significant main effect showed that those with a greater proclivity to use cannabis for health purposes had higher initial levels of use at baseline and faster increases in the rate of use over time. Regardless of risky DM, using cannabis for health purposes is associated with faster increases in cannabis use escalation. Additionally, risky DM does impact the association between recreational motives for use and cannabis use trajectories. Future work should examine these associations with additional motives for cannabis use that have been previously validated within the literature.

The positive feedback loop between academic self-efficacy, academic initiative, and Grade Point Average: a parallel process latent growth curve model

This study investigates the associations between students’ developmental changes in academic self-efficacy, academic initiative, and grade point average (GPA) during a three-year upper secondary education. The sample consisted of 1453 students aged 16–19 (60.6% girls; baseline mean age = 17.00, SD = .91; 56.1% high perceived family wealth; and 74.9% Norwegian-born). To explore how changes in academic self-efficacy, academic initiative, and GPA were related, we investigated a theoretical parallel process latent growth curve model. The results implied that, during upper secondary school, academic self-efficacy declined, while academic initiative and GPA remained stable. We found possible ceiling effects within and between several of the study’s constructs. The main finding was support for a positive feedback loop between the developmental trajectories of academic self-efficacy, academic initiative, and GPA. The present study adds new insight that should be taken into consideration when promoting positive educational development during late secondary school. Highlights Academic self-efficacy declines during upper secondary school, while academic initiative and grade point average remains stable. Parallel process latent growth curve model analysis. Possible ceiling effects within and between several of the study's factors. Positive associations between the trajectories of academic self-efficacy, academic initiative, and grade point average.

Latent Trajectories of Subjective Well-Being: An Application of Latent Growth Curve and Latent Class Growth Modeling

This study proposed a three-stage measurement model utilizing the Latent Growth Curve Modeling and Latent Class Growth Analysis. The measurement model was illustrated using repeated data collected through a four-week prospective study tracking the subjective well-being of volunteer college students (n=154). Firstly, several unconditional growth models were estimated to define the model providing a better representation of individual growth trajectories. Secondly, several conditional growth models were formulated to test the usefulness of covariate variables hypothesized to explain observed variance in growth factors. Finally, latent class models were estimated to explore different latent trajectory classes further. Results showed that students' subjective well-being changed over time, and the rate of this change, as well as its covariates, were not constant for the entire sample. This study clearly illustrates how a longitudinal measurement approach can enhance the scope of findings and the depth of inferences when repeated measurements are available.

Co-occurring Stress Trajectories and the Longitudinal Coupling of Internalizing Symptoms in Parent-Adolescent Dyads.

Stress is one candidate mechanism posited to contribute to the intergenerational risk of psychopathology. However, the ways in which parent and child stress are related across adolescence, and the role that co-occurring parent and child stress may exert regarding bidirectional risk for internalizing symptoms, are not well understood. Using repeated measures data spanning 3-years, this study investigated (1) the extent to which trajectories of parent and child stress are related during adolescence, and (2) whether co-occurring parent and child stress trajectories mediate prospective, bidirectional associations between parent depression symptoms and child internalizing symptoms (depression, physical and social anxiety). Participants included 618 parent-adolescent dyads (age 8-16; 57% girls; 89% mothers). Parent depressive symptoms and child symptoms of depression, social anxiety, and physical anxiety were assessed via self-report questionnaire at baseline and 36 months later. Parent and child stress were assessed via self-report questionnaire every three months between 3- and 33-months (11 total assessments). Latent growth curve model (LGCM) analysis found that parent and child stress trajectories were positively related across development. Prospective LGCM mediation analysis showed that higher youth stress at 3-months partially mediated prospective relations between parental depressive symptoms at baseline and youth depressive, as well as physical and social anxiety symptoms at 36-months. Parent and child stress reinforce each other across adolescence and may lead to increased risk for psychopathology. Increases in child stress represent an important factor conferring transdiagnostic risk for internalizing among children of depressed parents.

Predictive factors associated with trajectory of job satisfaction of workers with physical disabilities: A latent growth curve analysis.

Job satisfaction is a highly useful factor in maintaining a successful work life for the disabled. This study aimed to evaluate how job satisfaction among workers with physical disabilities evolves over time, and to identify predictive factors that influence the trajectory of job satisfaction. A longitudinal data analysis using the three-year databases of the 2016-2018 Employment Panel Survey for the Disabled (n = 693). A latent growth curve model analysis was conducted to determine the trajectory of the job satisfaction among workers with physical disabilities, as well as the predictive factors affecting it. Job satisfaction among workers with physical disabilities increased over time, and there were substantial individual differences in the trajectory of the job satisfaction. The significant predictors for the job satisfaction over the three years were degree of disability, computer usage ability, employment support services, and experience of discrimination. Gender, educational background, interpersonal relationship/organizational adaptability, convenience facilities for the disabled, job suitability, disability acceptability, and self-esteem had a considerable impact on the intercept of the job satisfaction, but not on the slope. The findings would inform healthcare providers to establish mid-to-long-term vocational rehabilitation intervention plans that could increase the job satisfaction among workers with physical disabilities.

How prepartum appearance-related attitudes influence body image and weight-control behaviors of pregnant Japanese women across pregnancy: Latent growth curve modeling analyses

A thin female physique is idealized in modernized societies, leading many pregnant women to experience body dissatisfaction and avoid weight gain, as pregnancy directs their body away from the sociocultural ideal of thinness. However, not all pregnant women report body dissatisfaction and/or weight-control behavior. We aimed to discern how prepartum weight-related attitudes—thin-ideal internalization, drive for thinness, and appearance comparison tendency—influence changes in negative attitudes toward a pregnant body, body image, and weight-control behavior across pregnancy. Data were longitudinally collected from 208 Japanese pregnant women (Mage = 27.0, SD = 1.74) across three pregnancy periods. Results indicated that overall, negative attitudes toward appearance change during pregnancy, weight-control behaviors, and restrained eating significantly changed across pregnancy. Moreover, those with higher prepartum thin-ideal internalization showed greater decreases in negative attitudes toward appearance change during pregnancy and greater increases in body dissatisfaction, whereas those with higher prepartum drive for thinness and appearance comparison tendency showed greater decreases in restrained eating across pregnancy. Our findings suggest that pregnant women with certain prepartum weight-related attitudes may experience an aggravation of body dissatisfaction while following dietary regimen during pregnancy. Thus, prevention programs may selectively help them reduce their negative body image during pregnancy.

The Birth Cohorts Difference in Depression.

Cohort effect refers to the social phenomenon that a certain event manifests differently depending on a group that is born in the same year or a similar time period. It is important to understand adolescents' depressive symptoms vary from generation to generation. We studied the changes of the depressive symptoms during the eighth to tenth grade and examined the difference between the 1997-birth cohort and 2000-birth cohort. The study included 2,070 students (2014-2016 of 2000-birth-cohort data, from 8th to 10th grade, mean age = 12.95-14.95 years) and 2,278 students (2011-2013 of 1997-birth-cohort data, from 8th to 10th grade, mean age = 12.90-14.90 years) who participated in the Korean Children and Youth Panel Survey. Latent growth curve modeling and multi-group analysis were used to compare depressive symptom trajectories between cohorts. The result of comparing the depression levels of the cohorts showed the initial values and change rates verifying the presence of a significant cohort effect-the depression level of the 2000-birth-cohort (initial value = 17.75) was significantly lower than the depression level of the 1997-birth-cohort (initial value = 19.49). This study has significance in that adolescents' depression has both age and cohort effects. It is speculated that the cohort effect may have been impacted by two major events: The Children Problem-Behavior Screening Questionare Test (mental health screening test), implemented nationwide in 2012, and the Sewol Ferry Disaster, which occurred in 2014. Based on the results, suggestions for future research were discussed.

Nighttime Sleep Quality and Daytime Sleepiness Predicts Suicide Risk in Adults Admitted to an Inpatient Psychiatric Hospital

ABSTRACT As sleep problems have been identified as an important, yet understudied, predictor of suicide risk, the present study analyzed the relationship between daytime sleepiness and nighttime sleep disturbance in a high-risk population of adults admitted to an inpatient psychiatric hospital. Objectives were to (1) examine the time course of subjective daytime sleepiness, nighttime sleep disturbance, and suicide risk throughout inpatient psychiatric treatment, (2) examine pre- to post-treatment changes in sleep disturbance with treatment as usual in an inpatient psychiatric setting, and (3) investigate whether daytime sleepiness and nighttime sleep disturbance predicted suicide risk above and beyond anxiety and depression. Participants were 500 consecutively admitted adults admitted to an intermediate length of stay (4–6 weeks) inpatient psychiatric hospital (47% female; 18–87 years of age). Measures of sleep, suicide risk, depression, and anxiety were completed at admission, weeks 1 through 4, and at discharge. Latent growth curve modeling (LGM) and hierarchal linear modeling (HLM) were conducted. The LGM analysis demonstrated that daytime sleepiness, nighttime sleep disturbance, and suicide risk all improved throughout inpatient treatment. Further, HLM showed that daytime sleepiness predicted suicide risk above and beyond symptoms of anxiety, depression, major sleep medications, and prior suicidal ideation and attempts, while nighttime sleep disturbance predicted suicide risk above and beyond symptoms of anxiety, major sleep medications, and prior suicidal ideation and attempts. Findings indicate the need to reevaluate safety protocols that may impact sleep, particularly that may increase daytime sleepiness, and to develop evidence-based sleep interventions for individuals admitted to inpatient psychiatric hospitals.

Social Disconnectedness and Loneliness: Do Self-Perceptions of Aging Play a Role?

Research suggests that self-perceptions of aging (SPA) have effects on physical, mental, cognitive, and emotional well-being among older adults. Few studies have examined the effects of SPA on social well-being. This study investigates the association of SPA with trajectories of social disconnectedness and loneliness in older Americans and explores mechanisms linking SPA and the 2 forms of social isolation. We conducted Latent Growth Curve Modeling and path analysis using 3-wave data spanning 8 years (2008/2010-2016/2018) from the Health and Retirement Study. The sample included respondents aged 65 and older (N = 3,597) at baseline (2008/2010). SPA was measured by the Attitudes Toward Own Aging Scale. Social disconnectedness was an index including 6 indicators of social networks and social engagement. Loneliness was measured using the 11-item UCLA Loneliness Scale. Older adults with more negative SPA at baseline were more lonely but had slower rates of increase in loneliness during the 8-year study period. More negative SPA also predicted greater social disconnectedness but was not significantly related to the rate of change in social disconnectedness over time. The effects of SPA on social disconnectedness were primarily indirect through loneliness, whereas SPA had direct effects on loneliness. Overall, SPA had a stronger association with loneliness than with social disconnectedness. The results suggest that older adults with negative SPA are at risk of loneliness which then leads to social disconnectedness. Asking about SPA during individual assessment with older adults may help to discern issues with loneliness.

Influences of target congruence and lifestyles on victimization trajectories: An Autoregressive Latent Trajectory Modeling approach to overlap between bullying and victimization

IntroductionRelative to the general victimization literature, a small body of literature has examined the longitudinal process of overlap between bullying perpetration and bullying victimization. To address this gap, this study assesses the dynamic process of bullying perpetration and victimization from a developmental perspective and examine the applicability of target congruence and lifestyle-routine activities approach. MethodsUsing data from 2,378 Korean youths collected from 2012 to 2016, the current study has conducted a Parallel Process Latent Growth Curve Modeling (PPLGCM) analysis to examine co-development and sequentially contingent processes of bullying and victimization, and Autoregressive Latent Trajectory Modeling (ALTM) to examine the time-lagged effect of bullying on victimization. ResultsResults from the PPLGCM revealed a significant and positive relationship between the initial levels of bullying and victimization as well as a reciprocal and cumulative influence over time. The ALTM analysis indicated that increased victimization in the previous year was associated with increased victimization in the following year. Further, time-concurrent and time-lagged effects of bullying on victimization were significant and positive. Finally, target congruence and risky lifestyle variables predicted the longitudinal process of victimization, indicating the cross-cultural generalizability of the theory. ConclusionsThe findings offer support for a complex and dynamic relationship between bullying and victimization, and thus call for developmentally sensitive prevention strategies.

The evolution of perfectionistic concerns and strivings in the proximity of exams: Evidence from a short-term longitudinal study

To date, little is known about the short-term dynamics of perfectionism in undergraduate students, especially in certain potentially activating contexts, such as the proximity of the exam session. The aim of the present study was to address this methodological limitation and literature gap by conducting a latent growth curve model analysis of students’ perfectionistic concerns and strivings in the proximity of exams, to detect and explain the potential heterogeneity in their growth trajectories. The study sample consisted of 242 undergraduate students who participated at four-time points (228 at T0, 171 at T1, 225 at T2, and 146 at T3), spaced four weeks apart, between the beginning of the first academic semester and the beginning of the exam session. The results showed that while perfectionistic concerns displayed a significant overall growth trajectory during this period of time, perfectionistic strivings were rather stable personality traits, less dependent on contextual factors. Regarding the individual differences in perfectionism dimensions, the findings showed that there was a significant inter-individual variability around both the initial level and the overall growth trajectory of perfectionistic concerns. However, only the perceived teachers' expectations explained significantly part of the variations around the initial level of perfectionistic concerns.

COVID-19 Perceived Impact and Psychological Variables as Predictors of Unhealthy Food and Alcohol Consumption Trajectories: The Role of Gender and Living with Children as Moderators.

The present study examines the trajectories of unhealthy food and alcohol consumption over time and considers whether perceived impact of COVID-19 and psychological variables are predictors of these trajectories. We ascertained whether these predictors are different in women vs. men and between women living with vs. without children. Data were collected through online surveys administered to 1038 participants from two universities (staff and students) in Chile, across five waves (July to October 2020). Participants provided information about their past-week unhealthy food and alcohol consumption and mental health. Using latent growth curve modeling analysis, we found that higher perceived health and interpersonal COVID-19 impact, younger age and lower depression symptoms were associated with more rapid increases over time in unhealthy food consumption. On the other hand, higher perceived COVID-19 economic impact and older age were associated with more rapid diachronic decreases in alcohol consumption. Gender and living with or without children, for women only, were moderators of these trajectories. This longitudinal study provides strong evidence identifying the multiple repercussions of COVID-19 and mental health factors on unhealthy food and alcohol consumption. These findings highlight the need for interventions aimed at minimizing the impact of the pandemic on unhealthy food and alcohol consumption over time.

System-justifying beliefs and trajectories of global self-worth among Black and Latinx college students.

The present study investigated the relationship between system-justifying beliefs (i.e., believing the United States to be fair and just for all racial and ethnic groups) during the 1st semester of college and trajectories of self-worth across 4 years of college among Black and Latinx college students attending an elite predominantly White institution. Participants in this 5-wave longitudinal study included 186 Black and Latinx college students (30% male; 26% 1st-generation college students; Mage = 18 years, SD = 0.36). The results of latent growth curve modeling analyses indicated that system-justifying beliefs were positively related to initial levels of self-worth and negatively related to trajectories of self-worth over time. Findings from the present study suggest that although initially beneficial, system-justifying beliefs may undermine self-evaluations among Black and Latinx college students over time. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

Longitudinal associations between positive affect and relationship quality among children and adolescents: Examining patterns of co-occurring change.

The broaden-and-build theory of positive emotions suggests that the experience of positive affect (PA) and positive interpersonal experiences should mutually reinforce each other over time, potentiating upward spirals of PA and social well-being. Informed by this upward spirals hypothesis, the present study used parallel process latent growth curve modeling to evaluate co-occurring trajectories of PA and parent- and peer-relationship quality in a sample of 680 youth aged 8 to 16 years (M = 11.87, SD = 2.41, 56.7% female). Youth PA and relationship quality were assessed every 18 months for 3 years (3 total assessments). Results of parallel process latent growth curve modeling analyses indicated positive correlations between trajectories of PA and trajectories of both parent- and peer-relationship quality. Consistent with tenets of the broaden-and-build theory of positive emotions, the findings suggest that youth PA and experiences of high-quality relationships demonstrate reciprocal patterns of growth, with implications for youth health and well-being. (PsycInfo Database Record (c) 2021 APA, all rights reserved).

A Comparative Study of Cohorts for Children’s Body Mass Index Changes

ABSTRACT The purpose of this study is to examine how the body mass index changes over time and to determine whether there is a difference between the cohorts. For these purposes, we used latent growth curve modeling and multi-group analysis involving 2,250 students (first-grade cohort panel data) and 2,094 students (fourth-grade cohort panel data) who participated in the Korean Children and Youth Panel Survey. Major findings are as follows. 1) The trajectories of body mass index were different by the cohort. Specifically, children’s body mass index has become more serious in recent generation. 2) Multi-group analysis showed that the trajectories of body mass index and its predictors were varied by the cohort. Specifically, self-esteem, abuse, sedentary lifestyle, and study time explained the cohort effect. These findings suggest that the social environment intervention is needed for the physical health of children.