Self-esteem trajectories over three decades predict opposition to social equality in midlife.

Although life trajectories are frequently theorized to explain people's attitudes toward different social groups, few studies have been able to directly assess their importance with suitable data. Addressing this gap and focusing on the development of general and domain-specific self-esteem, we report results from a population-based sample of Norwegians (N = 2,215) followed over 28 years and five time points from adolescence to midlife. Growth curve models demonstrated that irrespective of self-esteem domain, low levels of self-esteem in adolescence as well as a depressed self-esteem development over the next three decades were related to more overall opposition to social equality as well as more opposition to gender equality and immigration in midlife. The results held when controlling for participants' baseline political orientations and other key covariates in adolescence. Our findings indicate that low self-esteem and a lack of positive self-esteem development can be detrimental to harmonious intergroup relations in ever-diversifying societies. We discuss how future psychological interventions aimed at enhancing self-esteem may promote support for a more inclusive society.

Development of prosocial behavior and inhibitory control in late childhood: A longitudinal exploration of sex differences and reciprocal relations.

This study examined longitudinal development of prosocial behavior, assessed by the parent-reported Strength and Difficulty Questionnaire, and inhibitory control, measured by the Opposite Worlds Task, in a sample aged 9 and 12 years (n = 9468, 49.9% girls, 85.8% White) from the Avon Longitudinal Study of Parents and Children. The goal was to assess whether the level of prosocial behavior at age 9 relates to change in inhibitory control, and vice versa. Sex differences were also explored. Latent change score models showed that low inhibitory control in boys at age 9 was associated with more decreases in prosocial behavior from 9 to 12 years of age. This may suggest that interventions targeting inhibitory control in boys may also foster their social competence.

Secular trends in mental health problems among young people in Norway: a review and meta-analysis.

There is a growing concern that the mental health of recent generations of youth is deteriorating, yet the etiology of these secular changes is not fully understood. We aimed to review the evidence on trends in mental health problems among young people in Norway. Seven large-scale repeated cross-sectional studies were included in this study, comprising 35 cross-sectional data collections between 1992 and 2019, with a total sample of 776,606 young people. Our study found a clear increase in mental health problems among young females in Norway over the past few decades, while the trends were less marked for males. The proportion of individuals scoring above the problematic symptom score threshold increased on average by 11.2% (range 2.2% to 21.9%) for females and 5.2% (range - 0.9% to 11.1%) for males, based on data from the individual studies. The results from a meta-regression analysis showed that across all surveys, mean symptom scores increased by 17% (95% CI 12 to 21%) among females and 5% (95% CI 1 to 9%) among males from 1992 to 2019. Overall, mental health problems have increased continually since the early 1990s among young people, especially among young females. The cause of these secular changes remains unknown but likely reflect the interplay of several factors at the individual and societal level.Protocol registration: Open science framework, November 8, 2021 ( https://osf.io/g7w3v ).

Sexual and Physical Victimization and Health Correlates Among Norwegian Adolescents.

Large-scale epidemiological studies have documented that many children and adolescents are exposed to different forms of victimization experiences. However, such population-based studies have rarely examined how specific types of victimization are correlated with health indicators. Thus, we investigated sexual victimization, physical victimization by parents, and physical victimization by peers and their associations with sexual health, mental health, and substance use. We gathered data from a nationally representative sample of Norwegian 18-19-year-old students in their final year of senior high school (N = 2075; 59.1% girls). The analyses showed that 12.1% of the adolescents reported sexual victimization experiences. Physical victimization was more prevalent: 19.5% of the respondents had been exposed to victimization from parents and 18.9% from peers. Multivariate analyses revealed specific associations between sexual victimization and a range of sexual health indicators, such as early sexual intercourse debut, many sexual partners, engaging in sex without contraception while intoxicated, and participating in sexual acts for payment. Neither physical victimization from parents nor from peers were correlated with these variables. However, all three forms of victimization were associated with impaired mental health and potential substance use problems. We conclude that a variety of victimization experiences should be addressed in policies for prevention of adolescent mental health and substance use problems. In addition, a special emphasis is warranted regarding sexual victimization: Sexual health policies should address such potential experiences in addition to more traditional themes such as reproductive health and should also include low-threshold services for young victims of sexual victimization.

Sexual Victimization in the Digital Age: A Population-Based Study of Physical and Image-Based Sexual Abuse Among Adolescents.

Adolescents increasingly use social media platforms, and these practices open up new forms of sexual victimization, in particular image-based sexual abuse (IBSA). Few studies have examined prevalence rates and correlates of both physical sexual victimization (PSV) and these new forms of victimization in representative samples. We used data from 5,245 adolescent girls (53%) and 4,580 adolescent boys (47%) from the population-based Young in Oslo Study (mean age 17.1 years, SD = 0.9). Of all respondents, 2.9% had experienced IBSA, 4.3% PSV, and 1.7% both IBSA and PSV in the course of the previous 12 months. Multivariate analyses revealed that PSV victims, after control for other variables, had many characteristics described in previous studies of sexual victimization. Girls had higher prevalence rates than boys, many had been victims of other types of violence, and were part of peer groups with much use of alcohol and drugs. PSV victims also reported early intercourse onset and a higher proportion had been commercially sexually exploited. Lesbian, gay, and bisexual adolescents had higher victimization rates. Victims of both PSV and IBSA had a similar but even more pronounced profile. The IBSA victims were different: They lacked many of the traditional risk factors for sexual victimization, there were no significant gender differences in this group, and IBSA victims more often came from high socioeconomic backgrounds. In conclusion, we observe a reconfigured landscape of sexual victimization patterns among Norway adolescents due to their increasing participation on social media and digital platforms.

Childhood personality and academic performance: A sibling fixed-effects study.

OBJECTIVE: This study investigated the associations between personality traits at age 8 and academic performance between ages 10 and 14, controlling for family confounds. BACKGROUND: Many studies have shown links between children's personality traits and their school performance. However, we lack evidence on whether these associations remain after genetic and environmental confounders are accounted for. METHOD: Sibling data from the Norwegian Mother and Child Cohort Study (MoBa) were used (n = 9701). First, we estimated the overall associations between Big Five personality traits and academic performance, including literacy, numeracy, and foreign language. Second, we added sibling fixed effects to remove unmeasured confounders shared by siblings as well as rating bias. RESULTS: Openness to Experience (between-person ß = 0.22 [95% CI: 0.21-0.24]) and Conscientiousness (between-person ß = 0.18 [95% CI 0.16-0.20]) were most strongly related to educational performance. Agreeableness (between-person ß = 0.06 [95% CI -0.08-0.04]) and Extraversion (between-person ß = 0.02 [95% CI 0.00-0.04]) showed small associations with educational performance. Neuroticism had a moderate negative association (between-person ß = -0.14 [95% CI -0.15-0.11]). All associations between personality and performance were robust to confounding: the within-family estimates from sibling fixed-effects models overlapped with the between-person effects. Finally, childhood personality was equally predictive of educational performance across ages and genders. CONCLUSIONS: Although family background is influential for academic achievement, it does not confound associations with personality. Childhood personality traits reflect unbiased and consistent individual differences in educational potential.

Smoking in early adulthood is prospectively associated with prescriptions of antipsychotics, mood stabilizers, antidepressants and anxiolytics.

BACKGROUND: Whether smoking should be regarded as a risk factor for mental disorders remains unresolved. Prescribed psychotropic drugs can be used as indications for mental disorders. We investigated how smoking was prospectively related to prescription of antipsychotics, mood stabilizers, antidepressants, and anxiolytics. METHODS: Information about smoking, including the Fagerström Test for Nicotine Dependence, and relevant confounders, were obtained from the population-based Young in Norway Study (N = 2602), with four data collection waves between 1992 and 2006. These survey data were linked with information on prescriptions for psychotropic drugs from the comprehensive, nationwide Norwegian Prescription Database from 2007 to 2015. RESULTS: Daily smoking with high dependence in 2006 at age 28.5 (s.d. = 2.0) was associated with filling prescriptions of antipsychotics (OR, 6.57, 95% CI 2.19-19.70, p = 0.001), mood stabilizers (OR, 7.11, 95% CI 2.51-20.15, p < 0.001) and antidepressants (OR, 1.91, 95% CI 1.13-3.23, p = 0.016) 1-9 years later. Associations remained significant after adjustment for a variety of potential confounders measured before the assessment of smoking, including sociodemographic background, conduct problems, cannabis use, mental distress, and previous prescriptions for psychotropic medications. The association between smoking and prescription of anxiolytics was weaker and more unstable. CONCLUSIONS: In this study of young adults, daily smoking with high dependence was associated with later prescriptions of antipsychotics, mood stabilizers and antidepressants, indicating smoking as a risk factor for mental disorders treated with these drugs.

Prospective associations between loneliness and disordered eating from early adolescence to adulthood.

OBJECTIVE: Despite findings from numerous cross-sectional studies suggesting a substantial association between loneliness and different types of disordered eating, much remains unknown about the impact of confounding, the order of cause and effect, and gender differences in the relationship. Thus, this study followed a large, population-based, mixed-gender sample through adolescence and young adulthood, applying a random intercept cross-lagged panel model (RI-CLPM) approach to examine the bidirectional prospective associations between loneliness and disordered eating while ruling out the effect of unmeasured time-invariant confounders. METHOD: A Norwegian sample of N = 2933 adolescents (54.2% female) was examined across four time points (T1, Mage  = 15.44, grades 7-12; T2, Mage  = 16.93; T3, Mage  = 21.84; and T4, Mage  = 28.33) from 1992 to 2005 using RI-CLPMs for overall disordered eating and specific forms for disordered eating (dieting and bulimic symptoms). Multigroup structural equation models were used to assess gender differences. RESULTS: For male participants, high levels of loneliness at T1 predicted more overall disordered eating and more dieting at T2. Meanwhile, among female participants, disordered eating and bulimic symptoms at T2 predicted more loneliness at T3, whereas loneliness at T3 predicted more disordered eating and bulimic symptoms at T4, and vice versa. DISCUSSION: The findings suggest a pattern of bidirectional associations between loneliness and disordered eating that varies by time points, gender, and type of eating problem. Preventive interventions and treatment should consider social factors involved in the onset and maintenance of eating problems in male adolescents and young adult women. PUBLIC SIGNIFICANCE: This study contributes to the existing knowledge by examining for the first time the dynamic nature of the association between loneliness and disordered eating while accounting for all time-invariant confounding. Our findings reveal a pattern of bidirectional associations between loneliness and disordered eating that appears to vary by developmental period, gender, and type of eating problem. Our findings suggest that social factors have to be taken into account when designing prevention strategies aimed at disordered eating.

Posttraumatic growth during the COVID-19 lockdown: A large-scale population-based study among Norwegian adolescents.

The negative consequences of the COVID-19 lockdown during the spring of 2020 have been documented. However, adolescents may also have experienced positive personal and interrelational changes. This was the first study to examine the prevalence of posttraumatic growth (PTG) during the lockdown. We additionally explored how potential risk and protective factors, as well as experiences with the pandemic, were related to PTG and whether these associations were moderated by mental health resources and social support. We used data from a representative survey of 12,686 junior and senior high school students from Oslo, Norway, conducted during the lockdown (37% response rate, 56.4% girls). A short version of the Posttraumatic Growth Inventory was used to assess growth relative to personal strength, relationships with others, and appreciation of life. Several potential predictors in the domains of mental health, social relationships, experiences during the pandemic, and sociodemographic background factors were examined. Results from multiple regression analyses showed that satisfaction with life, parental care, worries about the pandemic, and immigrant status were the most prominent predictors of PTG, ßs = .14-.22, p < .001. Moderation analyses indicated a complex interplay between predictors of PTG by showing that good mental health was associated with higher degrees of PTG only in groups typically considered to be at higher risk of adverse outcomes. The findings provide information regarding who would profit from additional help to reinterpret the dramatic events during the lockdown to facilitate growth.

Basic human values and sick leave: A study combining two-wave survey data with longitudinal register data.

There is growing recognition that dispositional factors and social norms can predict work absence. Human values have been linked to related concepts, including work commitment and receipt of disability pension; however, there is a lack of research on whether human values are associated with sickness absence. We address this issue by combining survey data from two waves (2007, 2017) of the Norwegian Life-Course, Ageing and Generation Study (N = 1,330) with longitudinal register data on sickness absence between survey waves. Stepwise regression analyses showed that, out of Schwartz's 10 basic values, achievement was prospectively associated with higher levels of self-reported sick leave, even when controlling for a variety of potential confounders. Self-direction was also related to higher risk of self-reported sick leave in the adjusted analysis. Conservation values (security and conformity) were related to stricter attitudes toward sick leave when controlling for potential confounders, while stimulation was associated with lenient sick leave attitudes in the adjusted analysis. None of the human values were prospectively associated with longer-term register-based sick leave beyond bivariate correlations. We conclude that broad human values to some extent predict attitudes toward sick leave and self-certified sick leave where persons may vary according to which degree they consider sick leave to be necessary and appropriate, while human values do not predict long-term, physician-certified sickness absence. Future research may examine whether health- or work-specific values have greater explanatory power for sick leave, including long-term sickness absence that is typically more closely linked to more serious health problems.

Severe recalled labor pain and elective cesarean section in a subsequent delivery: a cohort study of Norwegian parous women.

INTRODUCTION: Some women keep on recalling intense labor pain experienced at childbirth to a degree that may negatively affect their life during the postpartum period or lead them to request a cesarean section (CS) in the subsequent delivery. This longitudinal study aimed to assess the impact of severe recalled labor pain from the previous birth on the preference for and delivery by an elective CS in the subsequent delivery. Furthermore, we investigated whether co-occurring maternal demographic, somatic and mental health factors related to the previous and subsequent delivery, explain parts of a potential association. MATERIAL AND METHODS: The study sample comprised 1135 parous women from the Akershus Birth Cohort. Severe recalled labor pain was assessed on a numeric rating scale at pregnancy week 17, and at pregnancy week 32, the preference for an elective CS for the subsequent delivery was assessed. Information on actual delivery by elective CS in the subsequent delivery was retrieved from the electronic birth record. Logistic regression analyses were conducted to examine the impact of severe recalled labor pain on elective CS. RESULTS: Severe recalled labor pain at the previous birth was associated with a preference for an elective CS (odds ratio [OR] 3.57, 95% confidence interval [CI] 2.25-5.67) and actual delivery by elective CS (OR 4.71, 95% CI 2.32-9.59). This association remained statistically significant for the preference for an elective CS (adjusted OR [aOR] 2.12, 95% CI 1.24-3.62) but diminished for delivery by elective CS (aOR 2.30, 95% CI 0.99-5.35) when adjusting for a variety of covariates. Factors related to previous childbirth such as number of years since previous birth, assisted vaginal delivery, anal sphincter lesions, overall birth experience and fear of childbirth were also linked to preference for and delivery by an elective CS. CONCLUSIONS: Women with severe recalled labor pain were about twice as likely to prefer an elective CS compared with women without severe recalled pain. For actual delivery, the significant association with severe recalled pain diminished after adjustment for covariates. However, sample size was small and, irrespective of severe recalled labor pain, preference for an elective CS was statistically significantly associated with actual delivery by elective CS.

Intelligence, alcohol consumption, and adverse consequences. A study of young Norwegian men.

AIMS: Research suggests that intelligence is positively related to alcohol consumption. However, some studies of people born around 1950, particularly from Sweden, have reported that higher intelligence is associated with lower consumption and fewer alcohol-related problems. We investigated the relationships between intelligence, alcohol consumption, and adverse consequences of drinking in young men from Norway (a neighboring Scandinavian country) born in the late 1970s. METHODS: This analysis was based on the population-based Young in Norway Longitudinal Study. Our sample included young men who had been followed from their mid-teens until their late 20s (n = 1126). Measures included self-reported alcohol consumption/intoxication, alcohol use disorders (AUDIT), and a scale measuring adverse consequences of drinking. Controls included family background, parental bonding, and parents' and peers' drinking. Intelligence test scores-scaled in 9 "stanines" (population mean of 5 and standard deviation of 2)-were taken from conscription assessment. RESULTS: Men with higher intelligence scores reported average drinking frequency and slightly fewer adverse consequences in their early 20s. In their late 20s, they reported more frequent drinking than men with lower intelligence scores (0.30 more occasions per week, per stanine, age adjusted; 95% CI: 0.12 to 0. 49). Intelligence was not associated with intoxication frequency at any age and did not moderate the relationships between drinking frequency and adverse consequences. CONCLUSIONS: Our results suggest that the relationship between intelligence and drinking frequency is age dependent. Discrepancies with earlier findings from Sweden may be driven by changes in drinking patterns.

Adolescents' assessments of consequences of the pandemic after one year of COVID-19 restrictions. / Ungdoms opplevelser av konsekvenser av pandemien etter ett år med covid-19-restriksjoner.

BACKGROUND: We have obtained knowledge of how the COVID-19 pandemic affected the lives of adolescents immediately after the pandemic arrived in Norway. However, we know little about adolescents' experiences from the pandemic over time, and whether sociodemographic factors and infection rates at the municipality level play a role. MATERIAL AND METHOD: We used questionnaire data from 106 448 lower and upper secondary school pupils who took part in the Ungdata survey in 167 municipalities in the spring of 2021 (response rate 76 %). The adolescents' responses regarding their experiences during the pandemic were collated with statistics on infection rates at the municipality level. We performed a Chi-square test and multilevel analyses to investigate predictors of adolescents' experiences. RESULTS: A total of 49 % responded that the COVID-19 pandemic affected their lives in a partly or very negative direction. Many reported negative changes in peer relationships, family relationships and mental health, but some also reported positive changes. Girls, older adolescents, those with a low socioeconomic background and those living in municipalities with a higher prevalence of infection reported more negative consequences. INTERPRETATION: Most adolescents reported that the pandemic has had more negative than positive consequences. Girls, older adolescents, those with a low socioeconomic background and those living in municipalities with a higher rate of infection may be especially affected by the negative effects of the pandemic.

Cannabis use in early adulthood is prospectively associated with prescriptions of antipsychotics, mood stabilizers, and antidepressants.

OBJECTIVE: Cannabis is an acknowledged risk factor for some mental disorders, but for others the evidence is inconclusive. Prescribed medicinal drugs can be used as proxies for mental disorders. In this study, we investigate how use of cannabis is prospectively related to prescription of antipsychotics, mood stabilizers, antidepressants, and anxiolytics. METHODS: Data on cannabis exposure and relevant confounders were obtained from 2,602 individuals in the longitudinal Young in Norway Study, providing survey data from four data collection waves between 1992 and 2006. Data were coupled with information about prescriptions for psychotropic drugs from the Norwegian Prescription Database between 2007 and 2015. RESULTS: Past year cannabis use increased the risk of prescription of antipsychotics (OR = 5.56, 95 % CI 1.64 - 18.87), mood stabilizers (OR = 5.36, 95 % CI 1.99 - 14.44) and antidepressants (OR = 2.10, 95 % CI 1.36 - 3.25), after accounting for sociodemographic variables, conduct problems, additional drug use, mental distress, and prescriptions the year before cannabis use was measured. CONCLUSIONS: In this study of young adults from the general population, past year cannabis use was associated with later prescriptions of antipsychotics, mood stabilizers, and antidepressants.

Posttraumatic stress symptoms following childbirth: associations with prenatal attachment in subsequent pregnancies.

This longitudinal population-based study aimed to investigate the prospective relationship between PTSD symptoms following childbirth and prenatal attachment in the subsequent pregnancy. Data were derived from the Norwegian Akershus Birth Cohort (ABC), a large population-based prospective cohort study. Data from 1473 women who had given birth at least once before and who had completed questionnaires at 17 and 32 weeks of gestation were included. Confirmatory factor analysis of the short version of the Prenatal Attachment Inventory was conducted to validate the scale. Further, structural equation modeling techniques were used to estimate prospective associations of PTSD symptoms following childbirth with prenatal attachment. Finally, to explore potential mechanisms of the association, mediation and moderation analyses were conducted. PTSD symptoms following previous childbirth were found to be prospectively related to higher levels of prenatal attachment in the subsequent pregnancy, while controlling for important confounding factors, such as symptoms of maternal depression and anxiety, previous pregnancy loss, and sociodemographic factors (maternal age, educational level, marital status, and number of children). When fear of childbirth was included as a potential mediating variable, the prospective relationship between PTSD symptoms following childbirth and prenatal attachment in the subsequent pregnancy increased, thereby indicating a suppressor effect. Fear of childbirth did not act as a significant moderator. Our findings suggest that a subsequent pregnancy following a traumatic childbirth may for some women represent an opportunity for a higher level of prenatal attachment, whereas high levels of fear of childbirth may be detrimental for prenatal attachment.

Longitudinal Pathways From Shyness in Early Childhood to Personality in Adolescence: Do Peers Matter?

Temperamental shyness in childhood is theorized to be an important contributor for adolescent personality. However, empirical evidence for such pathways is scarce. Using longitudinal data (N = 939 children, 51% boys) across 17 years, the aim of this study was to examine how shyness development throughout childhood predicted personality traits in adolescence, and the role of peers in these associations. Results from piecewise latent growth curve modeling showed early shyness levels to predict lower emotional stability and openness in adolescence, whereas early shyness levels and growth across childhood predicted lower extraversion. Peer problems in early adolescence accounted for these associations. This study is the first to demonstrate the role of childhood shyness and peer relations for adolescents' personality development.

Adolescent boys' physical fighting and adult life outcomes: Examining the interplay with intelligence.

Although it is well known that adolescent delinquent behavior is related to poor outcomes in adulthood, longitudinal research on specific acts of delinquency and their interplay with important individual characteristics in predicting future outcomes is scarce. We aimed to examine how physical fighting-one of the most common acts of violent delinquency among adolescent boys-is related to adult life success in several domains, and how intelligence influences these associations. The study used data from 1,083 boys that participated in the population-based longitudinal Young in Norway Study, following adolescents from 1992 to 2015, by combining self-reports at four time points with comprehensive information from registers. Results showed that adolescent boys' physical fighting was associated with poor adult outcomes in the domains of employment, education, and criminal behavior. Associations remained significant even after controlling for conduct problems in general-which isolated the effects of fighting from other delinquent acts-as well as from a variety of other potential confounders. Detailed analyses on the interplay of physical fighting and intelligence showed that some parts of the associations between adolescent boys' fighting and several adverse adult outcomes could be ascribed to lower intelligence among the fighters. Moreover, intelligence moderated the relationship between physical fighting and adult education. Adolescent fighting was not related to educational attainment among boys with high intelligence, whereas boys with lower intelligence experienced detrimental effects of adolescent fighting. The analyses show the importance of considering adolescent boys' physical fighting as a potential risk factor for future social marginalization.

Physical Fighting and Leisure Activities among Norwegian Adolescents-Investigating Co-occurring Changes from 2015 to 2018.

After many years of decline in violent behavior among adolescents in several Western countries, recent official statistics indicate a possible trend change. So far, knowledge on how this change is related to co-occurring changes in leisure time activities is limited. Using two cross-sectional surveys from Oslo, Norway, this study found substantial increases in the prevalence of physical fighting from 2015 (N = 23,381; 51.6% girls) to 2018 (N = 25,287; 50.8% girls) in junior and senior high school. The rise in fighting was related to co-occurring changes in several leisure activities, including increasing time spent unsupervised by adults, rising digital media use, and rising cannabis use. The study emphasizes the importance of considering leisure time activities when addressing adolescent misbehavior.

A longitudinal study of cannabis use increasing the use of asthma medication in young Norwegian adults.

BACKGROUND: A small number of studies have shown that the use of cannabis increases the risk of bronchial asthma. There is, however, a paucity of longitudinal studies which are able to control for known risk factors of bronchial asthma. METHODS: Survey data from a population-based longitudinal study encompassing 2602 young adults followed for 13 years were coupled with individual prescription data on asthma medication (ß2-adrenergic receptor agonists and glucocorticoids for inhalation) from the Norwegian national prescription database, which covers the entire Norwegian population. Current cannabis use, gender, age, years of education, body mass index (BMI; kg/m2) and current smoking were measured. RESULTS: Prescription of asthma medication was associated with female gender, self-reported earlier asthma and allergies, daily tobacco smoking and current cannabis use. In a model adjusting for gender, age, years of education, BMI, earlier self-reported asthma and allergies and current tobacco smoking the odds ratio for a current cannabis user to fill prescriptions for asthma medication was 1.71 (95% CI: 1.06-2.77; p = 0.028). CONCLUSIONS: This suggests that cannabis is a risk factor for bronchial asthma or use of asthma medication even when known risk factors are taken into consideration. Intake of cannabis through smoking should be avoided in persons at risk.

Predicting hazardous drinking in late adolescence/young adulthood from early and excessive adolescent drinking - a longitudinal cross-national study of Norwegian and Australian adolescents.

BACKGROUND: Research has consistently shown that early onset of drinking (EOD) is associated with alcohol-related problems in adulthood. However, recent reviews have identified several limitations in the early onset literature, including the use of retrospective reports, insufficient control for potential confounders, ambiguous definitions of the concept, and an assumption that early onset is independent of cultural norms and national alcohol policies. This study addresses these limitations by examining whether EOD, independent of early onset of excessive drinking (EOE), prospectively predicts hazardous drinking in late adolescence/young adulthood in Norway and Australia, two countries with different drinking cultures. METHODS: Data were drawn from two population-based longitudinal studies; the Norwegian Tracking Opportunities and Problems Study (n = 329) and the Australian International Youth Development Study (n = 786). Data were collected prospectively from mid adolescence (14-16 years) to late adolescence/young adulthood (18-25 years) and a modified Poisson regression approach was used to estimate prevalence ratios. Adolescent self-reports included measures of EOD and EOE. Young adults completed the Alcohol Use Disorders Identification Test (AUDIT). The results were adjusted for adolescent factors; age, gender, impulsivity, hyperactivity, conduct problems, smoking, early sexual intercourse and friends' substance use, and family factors; alcohol and drug use in the family, maternal education, family management and monitoring. RESULTS: Hazardous drinking was identified in 46.8 and 38.9% of young adults in Norway and Australia, respectively. Both EOD and EOE in adolescence were significantly related to an increased risk of alcohol-related problems in late adolescence/young adulthood in both studies, even when adjusting for possible confounders. CONCLUSION: Our findings indicate that adolescent drinking behaviour is an indicator of alcohol-related problems in late adolescence/young adulthood, even when controlling for a variety of covariates. This finding is in contrast to previous research on older adults, where no association between adolescent drinking and later alcohol-related problems were found when controlling for covariates. The divergence in findings may suggest that the impact of EOD/EOE is limited to the late adolescent and young adult period. Preventing drinking in early adolescence may thus have some impact on the drinking patterns in late adolescence/young adulthood.