Longitudinal Latent Class Analysis Research Articles

Middle Childhood Sport Participation Predicts Timely Long-Term Chances of Academic Success in Boys and Girls by Late Adolescence.

School age children experience successive academic demands which increase over time. Extracurricular sport develops skills which involve physical movement, social rules, formal practice, and rational competition. This may facilitate success. Using a prospective-longitudinal birth cohort of 746 girls/721 boys, we examined prospective associations between middle childhood sport participation with subsequent indicators of academic success in adolescence. We hypothesized that persistent participation would promote long-term achievement indicators. Mothers reported on whether the child participated in sporting activities with a coach/instructor from ages 6 to 10 years. Developmental trajectories of participation were generated using longitudinal latent class analysis and then used to predict outcomes. From ages 12 to 17 years, youth self-reported on academic indicators of success over the last 6 months. These were linearly regressed on trajectories of participation in sport, while controlling for pre-existing and concurrent child/family confounds. Children's sport participation from kindergarten to fourth grade predicted long-term chances of academic success in boys and girls, above and beyond individual/family confounders. Consistent participation predicted increments indicators of future success and reductions in academic failure and dropout risk through to the senior year of secondary school, ranging from 9.1% to 21.3% for girls and 11.7% to 22.9% for boys. We provide compelling and timely evidence of long-term associations between children's sport participation and subsequent indicators of academic success in typically developing boys and girls by late adolescence. Persistent sport involvement in sport increased educational prospects by late adolescence, potentially improving opportunities for success in emerging adulthood.

Risk of Unfavorable Trajectories of Chronic Pain Severity—Results of a Longitudinal Study in School Children

Chronic pain is a common burden among children and adolescents associated with impairments in many aspects of life. Higher pain severity increases this burden. Subsequently, it is important to identify factors predicting the course of pain severity, classified by the chronic pain grading (CPG). In a 1-year longitudinal assessment of a general school-age population (N = 2,280), we aimed to identify biopsychosocial factors associated with CPG trajectories. We focused on children and adolescents who reported chronic pain at the start of the year (N = 689). Using longitudinal latent class analysis, we identified 3 classes of CPG trajectories over 1 year: 1) the “pain recovery group” exhibited initially moderate CPG scores that rapidly and consistently declined to a pain-free level, 2) the “continuously moderate pain severity group” displayed initially high CPG levels with a mild decline over time, and 3) the “continuously low pain severity group” had initially moderate CPG levels, which only slightly declined. In comparison to the pain recovery group, the continuously moderate pain severity group presented with heightened levels of anxiety (odds ratio [OR] = 1.12, 95% confidence interval [CI] [1.02, 1.24], P = .023), depression (OR = 1.10, 95% CI [1.01, 1.19], P = .029), and affective pain perception (OR = 1.10, 95% CI [1.02, 1.18], P = .010) and were more likely to be female (OR = 2.14, 95% CI [1.05, 4.35], P = .036). The continuously low pain severity group was predominantly female (OR = 1.65, 95% CI [1.10, 2.49], P = .016) compared to the pain recovery group. In conclusion, girls and individuals with impaired psychological well-being more often exhibit unfavorable trajectories of chronic pain severity. PerspectivePediatric chronic pain patients, particularly females and those exhibiting elevated anxiety or depression scores or heightened affective pain perception warrant special attention in health care. These individuals have a greater risk of an unfavorable trajectory of chronic pain severity and might need more urgent and specialized treatment.

Abstract 4867: Longitudinal latent class analysis to further understand trajectory of density over time and risk of breast cancer

Abstract Background: It is clinically important to refine strategies to manage women with dense breasts as they represent approximately 50% of all women screened. Change in breast density is related to risk of breast cancer.1 We also must understand this change in density over time in women with dense and non-dense breasts. While population level data have been reported to estimate decrease in density with age, this has largely used digitized film images.2 Expanding use of digital mammography and repeated screening generates a large library of images for each woman, offering the potential to use more of the image data. Methods: We use longitudinal latent class analysis to cluster women and estimate their change in density over time. To evaluate the groups of women with different patterns of change in density we fit latent class models to our previously published data.1 This includes 289 pathology confirmed cases of breast cancer and 658 controls among women followed from November 3, 2008 to October 31, 2020. Results: Among women with dense breasts, there are 107 who developed breast cancer during follow-up and 267 control women who remained free from breast cancer. In the women without dense breasts, 173 cases and 383 controls were evaluated and showed similar age and BMI within the controls as seen in women with dense breasts. Prevalence of family history and history of benign breast biopsy did not differ between the two groups. We observe this latent class and trajectory of density phenomenal separately for dense and non-dense groups in our data. We show that the data define 2 classes, those who have decrease in density over time and those who have increase over time. The odds ratio (OR) for cancer comparing decline vs. incline is 4.9 in women with dense breasts (BI-RADS C, D) and 6.17 on women with non-dense breasts (BI-RADS A, B). Discussion: To refine our understanding of patterns of change in breast density over time in relation to breast cancer risk we fit latent class models. Longitudinal change matters as reflected in the 2-class model. This is independent of starting density and confirms that a one-time measure of density is not enough to define level of risk for subsequent breast cancer. Further work is needed to better define the drivers and inhibitors of decline in breast density over time given this change over time as women age is a universal phenomenon.2 1. Jiang S, Bennett DL, Rosner BA, Colditz GA. Longitudinal Analysis of Change in Mammographic Density in Each Breast and Its Association With Breast Cancer Risk. JAMA Oncol 2023;9(6):808-814. (In English). DOI: 10.1001/jamaoncol.2023.0434. 2. Burton A, Maskarinec G, Perez-Gomez B, et al. Mammographic density and ageing: A collaborative pooled analysis of cross-sectional data from 22 countries worldwide. PLoS Med 2017;14(6):e1002335. DOI: 10.1371/journal.pmed.1002335. Citation Format: Graham A. Colditz, Debbie L. Bennett, Shu Jiang. Longitudinal latent class analysis to further understand trajectory of density over time and risk of breast cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 4867.

Medicinal cannabis use among young adults during California’s transition from legalized medical use to adult-use: a longitudinal analysis

ABSTRACT Background: In 2016, California transitioned from legalized medical cannabis use to adult-use. Little is known about how this policy change affected medicinal cannabis use among young adults. Objectives: To identify longitudinal groups of medicinal cannabis users and concurrent changes in health- and cannabis use-related characteristics among young adults in Los Angeles between 2014 and 2021. Methods: Cannabis users (210 patients and 156 non-patients; 34% female; ages 18–26 at baseline) were surveyed annually across six waves. Longitudinal latent class analysis derived groups from two factors – cannabis patient status and self-reported medicinal use. Trajectories of health symptoms, cannabis use motives, and cannabis use (daily/near daily use, concentrate use, and problematic use) were estimated across groups. Results: Three longitudinal latent classes emerged: Recreational Users (39.3%) – low self-reported medicinal use and low-to-decreasing patient status; Recreational Patients (40.4%) – low self-reported medicinal use and high-to-decreasing patient status; Medicinal Patients (20.3%) – high self-reported medicinal use and high-to-decreasing patient status. At baseline, Medicinal Patients had higher levels of physical health symptoms and motives than recreational groups (p < .05); both patient groups reported higher level of daily/near daily and concentrate use (p < .01). Over time, mental health symptoms increased in recreational groups (p < .05) and problematic cannabis use increased among Recreational Patients (p < .01). Conclusions: During the transition to legalized adult-use, patterns of medicinal cannabis use varied among young adults. Clinicians should monitor increases in mental health symptoms and cannabis-related problems among young adults who report recreational – but not medicinal – cannabis use.

Cognition and Wealth Changes in Mid-to-later Life: A Latent Class Trajectories Approach Using the Health and Retirement Study.

ObjectivesTo assess how cognitive trajectories from mid-to-later life relate to wealth change, overall and by mid-life income. Methods: Data were from participants (51-64years) in the 2000-2018 U.S. Health and Retirement Study who were cognitively healthy at baseline (year 2000; unweighted n = 3821). Longitudinal latent class analyses generated cognitive and wealth trajectories, independently, and multinomial logistic regressions estimated the association between cognitive trajectories and wealth profiles, overall and by median income. Results: We identified three cognitive: cognitively healthy (CH), increasing cognitive impairment (ICI), and increasing dementia (ID) and four wealth profiles: stable wealth loss (SWL), delayed gradual wealth loss (DGWL), stable wealth gain (SWG), and gradual wealth gain (GWG). The ID group had higher probability of being in the SWL group and lower probability of SWG, which was more pronounced in respondents with greater median income. Discussion: Individuals with ID may be vulnerable to wealth loss, particularly for middle-class households.

Stability and shifts in the combined positive and negative mental health of clergy: A longitudinal latent class and latent transition analysis study of united methodist pastors before and after the onset of COVID-19

COVID-19 and its associated restrictions presented unprecedented challenges for those in the helping professions. In this study, we seek to understand how the mental health of those who belong to one specific helping profession – clergy – changed in the context of COVID-19. Using longitudinal data of a sample of United Methodist pastors from the North Carolina Clergy Health Initiative, we conduct both cross-sectional and person-centered analyses to investigate how the overall mental health of this occupational group changed, as well as how different subgroups of clergy fared within the context of the pandemic, depending on their well-being prior to the onset of COVID-19. We found that the mental health of pastors suffered within the context of the pandemic, but that individual changes in mental health differed based on what the combined positive and negative mental health patterns of clergy were prior to the pandemic, for which we used latent class analysis to identify as Flourishing, Distressed, Languishing, or Burdened but Fulfilled. Of these subgroups, having Flourishing pre-pandemic status was protective of mental health following the onset of COVID-19, whereas the other three subgroups’ mental health statuses worsened. This study is the one of the first longitudinal studies of helping professionals which has tracked changes in mental health before and after the onset of COVID-19. Our findings demonstrate the utility of considering positive and negative mental health indicators together, and they point to certain groups that can be targeted with well-being resources during future periods of acute or abnormal stress.

Identifying high-risk subgroups for self-harm in adolescents and young adults: A longitudinal latent class analysis of risk factors

BackgroundSelf-harm is a major public health concern in young people and is governed by a complex interaction of different risk factors. While many studies have identified these risk factors, less is known about how they tend to co-occur together. MethodsA latent class analysis was conducted using risk factors for self-harm from two waves at ages 13 and 17 from the Growing Up in Ireland (GUI) study their associations with self-harm were examined longitudinally at ages 17 and 20, respectively. ResultsAt age 13, there was a ‘peer problems’ group with higher probabilities of bullying, being unpopular and internalising problems and a ‘school and substance use problems’ group with higher probabilities of substance use, truancy or excessive absence from school and violence. Both of these two groups had over a two-fold risk of self-harm at age 17 in comparison to the low risk factor group. At age 17, there was a group with depression that was diagnosed by a medical professional with the highest relative risk (RR:13.9 (95 % CI 10.2–19.0)) of self-harm at age 20. Two other groups with undiagnosed depression had high probabilities of being bullied, and either high or low probabilities of substance use that had a 9.4 (95 % CI 6.8–13.1) and 7.4 (95 % CI 5.5–10.0) relative risk of self-harm at age 20, respectively. ConclusionsIdentifying hidden sub-groups using risk factors for self-harm in young people can inform potential public health interventions by clinicians and other professionals who work with young people.

Life-Course Associations Between Smoking and Depressive Symptoms. A 30-Year Finnish Follow-up Study.

Relatively little is known about whether the association between smoking and depressive symptoms changes with age and how the trajectories of smoking and depressive symptoms are intertwined during the life course. In this population-based study, these associations were examined from young adulthood to middle age. Participants of a Finnish cohort study (N = 1955) were assessed at the ages of 22, 32, 42, and 52 using questionnaires covering daily smoking (yes/no) and the short 13-item Beck Depression Inventory. Longitudinal latent class and longitudinal latent profile analyses were used to identify life course trajectories of smoking and depressive symptoms. The proportions of daily smokers decreased, while levels of depressive symptoms increased among both females and males from age 22 to 52 years. Smoking was associated with higher levels of depressive symptoms from age 22 to 42 years, while not at 52. Associations among males prevailed when adjusting for education, marital status, and alcohol use. Four life course classes of daily smoking (nonsmokers, decreasing prevalence of smoking, persistent smokers, and increasing prevalence of smoking) and four trajectories of depressive symptoms (low, increasing/moderate, decreasing/moderate, and high) were identified. In males, persistent daily smokers (relative risk ratio (RRR) = 4.5, 95% confidence interval (CI): 2.2 to 9.2) and those in the class with increasing smoking prevalence (RRR = 3.2, 95% CI: 1.1 to 9.1) had an increased risk of belonging to the high depressive symptoms profile. In females these associations were nonsignificant. Compared to females, the relationship between smoking and depressive symptoms seems more robust among males during adulthood. Specifically, males smoking persistently from young adulthood to middle age have an increased risk of high depressive symptoms trajectory. This population-based cohort with 30 years of follow-up showed that the life course trajectories of daily smoking and depressive symptoms are associated. Persistent daily smokers and those starting late had an increased risk of belonging to the profile with constantly high levels of depressive symptoms during the life course. However, these associations were statistically significant only in males. Actions should be strengthened, especially in males, to prevent smoking initiation, to help smoking cessation, and to identify and treat depression in smokers with significant depressive symptoms.

Opioid therapy trajectories of patients with chronic non-cancer pain over 1 year of follow-up after initiation of short-acting opioid formulations.

This study compared opioid utilization trajectories of persons initiating tramadol, short-acting hydrocodone, or short-acting oxycodone, and it characterized opioid dose trajectories and type of opioid in persistent opioid therapy subsamples. A retrospective cohort study of adults with chronic non-cancer pain who were initiating opioid therapy was conducted with the IQVIA PharMetrics® Plus for Academics data (2008-2018). Continuous enrollment was required for 6 months before ("baseline") and 12 months after ("follow-up") the first opioid prescription ("index date"). Opioid therapy measures were assessed every 7 days over follow-up. Group-based trajectory modeling (GBTM) was used to identify trajectories for any opioid and total morphine milligram equivalent measures, and longitudinal latent class analysis was used for opioid therapy type. A total of 40276 tramadol, 141023 hydrocodone, and 45221 oxycodone initiators were included. GBTM on any opioid therapy identified 3 latent trajectories: early discontinuers (tramadol 39.0%, hydrocodone 54.1%, oxycodone 61.4%), late discontinuers (tramadol 37.9%, hydrocodone 39.4%, oxycodone 33.3%), and persistent therapy (tramadol 6.7%, hydrocodone 6.5%, oxycodone 5.3%). An additional fourth trajectory, intermittent therapy (tramadol 16.4%), was identified for tramadol initiators. Of those on persistent therapy, 2687 individuals were on persistent therapy with tramadol, 9169 with hydrocodone, and 2377 with oxycodone. GBTM on opioid dose resulted in 6 similar trajectory groups in each persistent therapy group. Longitudinal latent class analysis on opioid therapy type identified 6 latent classes for tramadol and oxycodone and 7 classes for hydrocodone. Opioid therapy patterns meaningfully differed by the initial opioid prescribed, notably the presence of intermittent therapy among tramadol initiators and higher morphine milligram equivalents and prescribing of long-acting opioids among oxycodone initiators.

LONGITUDINAL ASSOCIATION OF COGNITIVE IMPAIRMENTS WITH LATE-LIFE WEALTH CHANGES AMONG OLDER AMERICANS

Abstract Disparities in wealth in the US are linked to disadvantageous health outcomes over the lifecourse. Cognitive impairment can widen wealth disparities. We examine how cognitive impairment influences change in wealth starting in mid-life and into older adulthood and the role of social determinants of health in wealth loss due to cognitive impairment. We use biennial longitudinal data (Health and Retirement Study: 2000-2018) on Non-Hispanic White (NH-W), NH-Black (NH-B), and Hispanic (H) middle-aged adults (51-64 years and cognitively healthy in 2000; unweighted n=3,651) to model wealth change (measured as the relative (%) difference in reported wealth between 2000-2018 [in 2018 dollars]) as a function of cognitive status (using three latent phenotypes derived from longitudinal latent class analyses: cognitively healthy [CH; 77%], increasing cognitive impairment [ICI; 19%], and increasing dementia [ID; 4%]). We fit racial/ethnic stratified regression models to examine relationships between cognitive phenotypes and wealth change, adjusting for covariates. Using Blinder-Oaxaca decomposition techniques, we examined the contributions of predisposing (e.g. race/ethnicity), health enabling (e.g. insurance source), and health needs (i.e. chronic conditions) factors in wealth change by cognitive phenotypes. Results show NH-Ws had higher levels of reported wealth in both 2000 and 2018, compared to NH-Bs and Hs. However, the relative rate of change in wealth was not-significantly different across race/ethnic groups. Education, insurance, and health explained most differences in wealth change related to cognitive impairment. Yet, nearly 50% of group differences in wealth changes remained unexplained. Future work should further elucidate sources of these unexplained differences, overall and by race/ethnicity.

Perceived Epidemic Impacts and Mental Symptom Trajectories in Adolescents Back to School After COVID-19 Restriction: A Longitudinal Latent Class Analysis

PurposeThis study aimed to assess the impacts of COVID-19 epidemic on various life aspects and identify the trajectories of common mental symptoms among adolescents back to school after COVID-19 restriction. Furthermore, potential predictors associated with those trajectories were investigated. MethodsThis longitudinal study, with five data collection points and a total follow-up of 68.4 days, was conducted among 1,393 junior high school students (mean age: 13.8 years; male, 53.3%) shortly after school reopened during the first COVID-19 outbreak in China. Questions on sociodemographics and perceived COVID-19 epidemic impacts were completed at the baseline while the Patient Health Questionnaire, Generalized Anxiety Disorder Scale, and Insomnia Severity Index were measured throughout the study for depression, anxiety, and insomnia symptoms, respectively. Trajectories of mental symptoms were classified by longitudinal latent class analysis, and the associated predictive factors were identified with multinomial regression modelling. ResultsOur study revealed high but steadily declining prevalence of depression, anxiety, and insomnia symptoms (p trend < .001). Five distinctive trajectories were identified for both depression and anxiety (“resistance,” “low symptom,” “recovery,” “chronic dysfunction,” and “delayed dysfunction”) and three for insomnia (“resistance,” “low symptom,” and “chronic dysfunction”). Besides the significant association between the mental symptom trajectories and students' perceived COVID-19 impacts on study practice, family income, and family relationship, female gender, lower school grade, and higher body mass index were found to be predictive of high severity trajectories. DiscussionOur findings may help locate the most psychologically vulnerable adolescents during the epidemic and foster better implementation of targeted intervention.

Change and stability in British drinking practices and culture between 2009 and 2019: A longitudinal latent class analysis of drinking occasions

RationaleTheories of practice can support understanding of health-related behaviours, but few studies use quantitative methods to understand time-trends in practices. This paper describes changes in the prevalence and performance of alcohol drinking practices in Great Britain between 2009 and 2019. MethodsLatent class analyses of annual cross-sectional data collected between 2009 and 2019. The dataset come from a one-week retrospective diary survey of adults resident in Great Britain. It contains 604,578 drinking occasions reported by 213,470 adults (18+) who consumed alcohol in the diary-week. The measures describe occasion characteristics including companions, location, motivation, timings, accompanying activities and alcohol consumed. We estimate separate latent class models for each year and for off-trade only (e.g. home), on-trade only (e.g. bar) and mixed-trade occasions. ResultsWe identified fifteen practices; four off-trade only, eight on-trade only and three mixed-trade. The prevalence of practices was largely stable over time except for shifts away from drinking with a partner and towards drinking alone in the off-trade, and shifts away from Big nights out and towards other forms of heavy drinking in the on-trade. We identified five key trends in the performance of practices: (i) spirits increasingly replaced wine as the main beverage consumed in occasions; (ii) home-drinking moved away from routinised wine-drinking with meals on weekdays and towards spirits-drinking on weekends; (iii) the Male friends at the pub practice changed less than other pub-drinking practices; (iv) Big nights out started later, often in nightclubs, and involved less pub-drinking or heavy drinking and (v) the meal-based and Going out with partner practice formats showed few changes over time. ConclusionKey recent trends in British drinking practices include a decline in routinised wine-drinking at home, a transformation of big nights out and a mixture of stability and change in pub- and meal-based practices.

Adolescent Substance Use Patterns and Risk for Suicidal Thoughts and Behaviors in Young Adulthood.

There is strong evidence that substance use is a risk factor for suicidality. Prior studies have focused on the suicide risk associated with using individual substances, even though substance users often use more than one substance. This study investigates the association between patterns of adolescent substance use and suicidality in young adulthood. Participants were U.S. adolescents (n = 2,111, 58.9% female, mean age = 16.31 years) from the NEXT Generation Health Study, which followed tenth graders for 7 years (2009/2010-2016) and collected data via yearly surveys. Longitudinal latent class analysis was used to identify high school patterns of substance use, and logistic regression was used to relate these patterns to risk of suicidality in young adulthood. We identified two groups of adolescents: "non-/infrequent users" (71%) and "multiple substance users" (29%). Multiple substance users had higher odds of making a suicide plan or attempt in young adulthood (odds ratio [OR] = 2.41, 95% CI [1.05, 5.53]), but not suicidal ideation (OR = 1.55, 95% CI [0.80, 2.99]), than non-/infrequent users, adjusting for sociodemographic factors and depressive symptoms. Multiple substance users with suicidal ideation were not more likely to progress to a later plan or attempt (OR = 1.41, 95% CI [0.41, 4.84]) than non-/infrequent users. Adolescents who use substances in high school are at higher risk for making a suicide plan or attempt in young adulthood. Early identification of these adolescents may help inform interventions to mitigate risk for suicidal behaviors in young adulthood.

Quality of life in elderly patients with venous thromboembolism assessed using patient-reported outcome measures

BackgroundWe aimed to evaluate the quality of life (QoL), using patient-reported outcome measures (PROMs), in elderly patients with venous thromboembolism (VTE) and to explore whether VTE complications (recurrence, bleeding, or postthrombotic syndrome) had an impact on later QoL. MethodsWe used data from the SWIss venous Thromboembolism COhort of older patients(SWITCO65+), a prospective multicenter cohort of patients aged ≥65 years with acute, symptomatic VTE. Primary outcome was changes in QoL up to 24 months, assessed using generic (36-Item Short-Form Health Survey), with physical (PCS) and mental component score (MCS), and disease-specific (Venous Insufficiency Epidemiological and Economic Study [VEINES]-QoL, [VEINES-Sym], and Pulmonary Embolism QoL) PROMs. PROM scores ranged from 0 to 100 points, higher scores indicating a better QoL. Longitudinal latent class analysis was used to group patients with similar PCS trajectories. Repeated-measures linear regression analyses were used to assess effects of VTE complications on changes in QoL scores. ResultsIn 923 patients (median age, 75; male, 54%), 140 (15%) patients died, 97 (11%) experienced recurrent VTE, and 106 (12%) major bleeding during follow-up. Compared with patients with higher PCS trajectories, patients with lower PCS trajectories were more likely to be older, female, sicker, and less physically active. On average, generic and disease-specific QoL scores improved over time (+11% in PCS, +3% in MCS, +6% in VEINES QoL, and +16% in Pulmonary Embolism QoL at 3 months). VTE complications were always associated with significantly lower QoL scores (for VTE recurrence: PCS adjusted difference -2.57, 95% CI, −4.47 to −0.67). ConclusionAlthough QoL following VTE tended to improve over time, patients with VTE-related complications had lower QoL than patients without complications.

Longitudinal Patterns of Multidimensional Violence Exposure and Adolescent Early Sexual Initiation.

Early sexual intercourse is associated with sexually transmitted infections, pregnancy, and depressive symptoms, and delay of intercourse allows adolescents opportunities to practice relationship skills (Coker et al., 1994; Harden, 2012; Kugler et al., 2017; Spriggs & Halpern, 2008). Thus, understanding predictors of early sexual intercourse is crucial. Prior research has suggested that violence exposure is associated with early initiation of sexual intercourse in adolescence (Abajobir et al., 2018; Orihuela et al., 2020). However, most studies have looked only at a single type of violence exposure. In addition, little research has examined longitudinal patterns of violence exposure in order to determine whether there are particular periods when the violence exposure may have the strongest impact on sexual behavior. Guided by life history and cumulative disadvantage theories, we use longitudinal latent class analysis and data from the Future of Families and Child Well-being Study (N = 3,396; 51.1% female, 48.9% male) to examine how longitudinal patterns of multiple types of violence exposures across ages 3 to 15 are associated with early sexual initiation in adolescence. Findings suggest that experiencing persistent physical and emotional abuse across childhood was associated with the greatest prevalence of early sexual initiation. Early exposure to violence was not consistently associated with greater likelihood of sexual initiation; instead, early abuse was more strongly associated with sexual initiation for boys, while late childhood abuse was more strongly associated for girls. These findings suggest that gender-sensitive programs are highly needed to address unique risk factors for boys' and girls' sexual behaviors.

Modeling Early Heterogeneous Rates of Progression in Boys with Duchenne Muscular Dystrophy.

Duchenne muscular dystrophy (DMD) exhibits substantial variability in rates of disease progression and response to treatment. This has hindered treatment development and complicated interpretation of drug effects in clinical trials. We hypothesized that a multivariate combination of early-age clinical outcome measurements can explain differential disease progression. Data on boys with DMD (ages 4-<10 years), both treated with steroidal anti-inflammatories and untreated, were obtained from CINRG Duchenne Natural History Study (n = 209) and vamorolone VBP15-002/003/LTE (n = 46) studies. Velocities from three timed function tests (TFTs; stand from supine, run/walk 10 meters, and climb 4 stairs) were simultaneously modeled in a longitudinal latent class analysis. Three classes of differentially progressing early age DMD motor trajectories were identified. Quicker decline/progression was associated with lower baseline TFT velocities, earlier loss of ability to finish a TFT, and lower predicted velocities. Earlier substantial steroid exposure was associated with greater TFT velocities while the moderate progression class was observed to have the largest difference in performance between boys treated early with steroids vs. not. Sample size calculations with the class showing the largest treatment response showed a large reduction in required sample size as compared to using summaries from all participants. Gene mutations were also investigated in post-hoc analyses, with mutations near the beginning of the DMD gene (Dp427 absent and Dp140/Dp71 present) found to be enriched in the slowest progressing class. This study provides insight into the variation in DMD progression through a latent class analysis. Our findings show class-related trajectories of motor outcomes and pharmacological response to corticosteroids, and suggest that enrichment strategies and/or subgroup analyses could be considered further in design of therapeutic interventions in DMD.

Development of well-being after moving to telework: A longitudinal latent class analysis.

Due to the COVID-19 pandemic, teleworking suddenly became a reality for many individuals. Previous research shows that there are contradictory consequences of telework on well-being: while workers have the opportunity for self-directed work, intensified work behavior as well as longer hours being worked might occur at the same time. We expect that the effects of telework vary over time and may be able to explain these contradictions. Moreover, from the perspective of the job demands-resources model besides job resources, personal resources may be relevant. The aim of this study is to investigate how the mental well-being of workers unfolds over time after the onset of the pandemic and the role of telework in this process. Additionally we seek to identify the impact of available job resources and personal resources in this extraordinary situation. Data were collected online from 642 participants in Germany beginning in March 2020, with 8 weekly followup surveys. Mental well-being was measured using the WHO-5 well-being index. For personal resources we looked at occupational self-efficacy; job resources were flexible working hours, job autonomy, and social support. Job demands were telework and work intensification. First we used a group-based trajectory analysis approach to identify different well-being trajectories. Second we applied multinomial regression analysis to identify T1 predictors of well-being trajectory group membership and their interactions. We found three groups of mental well-being trajectories: low, medium, and high. Their progress through the investigation period was rather stable: we observed only slight improvements of mental well-being for the high well-being group and a slight deterioration for the other two groups. Only the job demand work intensification and the personal resource occupational self-efficacy had a significant relationship to group assignment. Additionally we found interactions of telework with work intensification and occupational self-efficacy indicating a buffering mechanism of telework on the consequences of high work intensification; and low occupational self-efficacy. Telework appears to be a useful resource that buffered high work intensification and compensated for low personal resources during the pandemic. Since data were from self-reports of a convenience sample we can't assume generalization of our results nor absence of common-method bias.

Abstract P359: Long-Term Exposure to Neighborhood Poverty Throughout Midlife and Subclinical Cardiovascular Disease Burden in Subsequent Years: The Study of Women's Health Across the Nation

Purpose: To examine whether longitudinal exposure to neighborhood poverty throughout midlife influences the degree of subclinical CVD (SCVD) burden experienced in early old age in a racially and ethnically-diverse cohort of women transitioning through menopause. Methods: We used longitudinal data on 1,212 women (aged 42-52 at baseline) from The Study of Women’s Health Across the Nation . Residential address histories were collected annually for eleven years across midlife (1996-2007). The following carotid ultrasound SCVD measures were collected during a subsequent follow-up visit (2009-2013): intima media thickness (CIMT), inter-adventitial diameter (IAD), and plaque presence. Values &gt;75th percentile of each were considered high (CIMT &gt; 0.79 mm; IAD &gt; 7.21 mm; plaque index &gt;=2) and were summed together to create our composite SCVD index (range: 0-3). We used longitudinal latent class analysis to identify distinct patterns of exposure to high neighborhood poverty among women across midlife, and used ordinal logistic regression models to determine if a woman’s neighborhood poverty class throughout midlife was associated with her future SCVD burden. Results: We identified five longitudinal neighborhood poverty classes across midlife: consistent low (72% of total), decreasing (6.4%), early-increasing (7.3%), late-increasing (5.5%), and consistent high (8.4%). In age and site-adjusted models, the odds of having greater SCVD burden were 70% higher among women living in consistently high poverty neighborhoods (OR=1.70, 95% CI: 1.14-2.54), 52% higher among women experiencing early increases in neighborhood poverty (OR=1.52, 95% CI: 0.97-2.38), and 69% higher among women experiencing late increases in neighborhood poverty (OR=1.69, 95% CI: 1.04-2.74), compared to women living in consistent low poverty throughout midlife. Controlling for social, behavioral, and menopause-associated factors reduced the magnitude and significance of the estimated effects towards the null. Adjusting for baseline age, race/ethnicity, site, education, marital status, employment, alcohol use, smoking, depressive symptoms, physical activity, menopausal status, hypertension, diabetes, and obesity, the estimated odds of greater SCVD burden were 6% higher among women living in consistently high poverty neighborhoods (95% CI: 0.69-1.63), 10% higher among women experiencing early increases in poverty (95% CI: 0.68-1.76), and 17% higher among women experiencing late increases in poverty (95% CI: 0.70-1.94), relative to women living in consistent low poverty neighborhoods across midlife. Conclusions: Women living in neighborhoods with higher levels of poverty throughout midlife had greater SCVD burden in early old age. Associations between neighborhood poverty trajectories and SCVD burden were largely accounted for by individual-level social, behavioral, and biomedical confounders.

Children’s internalizing behavior development is heterogeneously associated with the pace of epigenetic aging

BackgroundInternalizing behaviors are an indicator of children’s psychological and emotional development, predicting future mental disorders. Recent studies have identified associations between DNA methylation (DNAm) and internalizing behaviors. This prospective study aimed at exploring the associations between pace of biological aging and the developmental trajectories of internalizing behaviors. MethodsParticipants were children from the Avon Longitudinal Study of Parents and Children (ALSPAC) cohort (N = 974). Measures of DNA methylation were collected at birth, age 7 and ages 15–17. The pace of aging was estimated using the DunedinPoAm algorithm (PoAm). Internalizing behaviors reported by caregivers between ages 4 and 16 using the Strengths and Difficulties Questionnaire. To explore heterogeneity in the association between PoAm and internalizing behaviors we use Poisson quantile regression in cross-section heterogeneity and longitudinal latent class analysis over the childhood and adolescence. ResultsInternalizing behavior trajectories were identified: low-risk, childhood limited, late onset and early onset (persistent). Accelerated aging at birth was negatively associated with internalizing behaviors in early childhood but positively correlated during adolescence. Higher PoAm at birth increased chance of low-risk profile, while decreasing likelihood of childhood limited trajectory. PoAm at age 15 was negatively associated with childhood limited profile and positively linked to late onset trajectories. Associations were larger at higher values of internalizing symptoms. ConclusionsThe heterogeneity in the association between biological age acceleration and internalizing behaviors suggests a complex dynamic relationship, particularly in children with high or increased risk of adverse mental health outcomes.

Parental monitoring, family conflict, and adolescent alcohol use: A longitudinal latent class analysis.

This study examined relationships among parental monitoring, family conflict, and subgroups of adolescent alcohol use identified through longitudinal latent class analyses (LLCA). Differences in these subgroups across sex and race/ethnicity were also examined. The present study used data (N = 4,067; 51% male, 49% female) collected during a longitudinal study in which adolescents completed questionnaires each semester for seven semesters, beginning in spring of their freshman year of high school until spring of their senior year. LLCA demonstrated three classes of drinking over time (low, increasing, and moderate use). The majority of adolescents were increasing-use drinkers. Moderate drinkers were more likely to be male than female; nondrinkers were most likely to be African American; and increasing-use drinkers were more likely to be Mexican American. Adolescents who received less maternal monitoring and experienced more family conflict were more likely to be identified in the class of moderate alcohol use compared to nonuse. These results highlight the importance of encouraging parental monitoring and decreasing family conflict to reduce the likelihood of adolescent alcohol use throughout the high school years (PsycInfo Database Record (c) 2022 APA, all rights reserved).