Recent research has identified genes linked to educational attainment, but their effects on subsequent socioeconomic outcomes, particularly in egalitarian Nordic welfare states, remain largely unexplored. We analyze two genetically informed Finnish datasets, encompassing longitudinal register information on earnings, employment, unemployment, occupational status, and social assistance receipt (n = 31,622). We examine the role of a polygenic score for educational attainment (EA PGS), achieved level of education, and family socioeconomic background in predicting these outcomes in adulthood. We further study cohort differences around Finland's comprehensive school reform of the 1970s that aimed to promote equality of opportunity. Our results show that in the post-reform generation, EA PGS did not significantly predict adulthood outcomes after controlling for the achieved level of education. A notable exception was for occupational status. In contrast, in the pre-reform generation, EA PGS predicted later socioeconomic outcomes beyond education, indicating relationships not fully explained by schooling. Parental income did not moderate the effect of the EA PGS. Our findings shed additional light on the mechanisms connecting genetic factors and life chances, demonstrating that institutional setting and schooling can shape the influence of genetic endowment for high educational attainment in adult socioeconomic status.

Importance:Peer networks may influence early psychiatric vulnerability, both through direct exposure to peers' mental disorders and via peers' genetic predispositions. Clarifying these peer influences could inform preventive strategies aiming to improve adolescent mental health. Objective:To determine whether adolescents' exposure to peers' diagnosed mental disorders or genetic predispositions is associated with subsequent risk of mental disorders and whether these associations differ across disorders or peer network contexts. Design, Setting, and Participants:Nationwide register-based cohort study including Finnish residents born between 1985 and 2000. Cohort members were assigned to 4 peer network contexts: lower secondary school, upper secondary school, postal code area, and 1000-m square. Follow-up extended from age 17 years until first mental disorder diagnosis, emigration, death, or December 31, 2023. Study data were analyzed from October 2025 to March 2026. Exposures:(1) Peers' family-based genetic risk scores (FGRS) for mental disorders, estimated from diagnoses in first- through fifth-degree relatives and (2) peers' own diagnoses of mental disorders. Main Outcomes and Measures:Time to an incident diagnosis of the disorder after age 17 years. Associations were estimated using Cox proportional hazards models adjusted for sex, birth year, proband FGRS, parental education and income level, and peer network size. Results:Of 604 819 cohort members (mean [SD] age, 17 [0] years; 303 967 male [50.3%]), 234 117 received a mental disorder diagnosis over a median (IQR) follow-up of 11.7 (7.5-16.5) years. Peer FGRS was associated with subsequent risk of the same disorder across disorders and peer networks, with the greatest associations for externalizing disorders in upper secondary school (hazard ratio [HR], 1.34; 95% CI, 1.29-1.38). In contrast, peer diagnoses showed the strongest associations for internalizing disorders in upper secondary school (HR, 1.17; 95% CI, 1.15-1.18). Cross-disorder associations were also observed, with peer FGRS for externalizing disorders associated with proband internalizing outcomes, and vice versa. However, whereas peer internalizing diagnoses were associated with both proband internalizing and externalizing outcomes, associations between peer externalizing diagnoses and proband internalizing outcomes were substantially weaker than those with proband externalizing outcomes. Conclusions and Relevance:Findings of this cohort study suggest that peer environments, particularly in upper secondary school settings, are associated with mental health trajectories. More research is needed to elucidate the mechanisms underlying these findings.

We investigate whether childhood adiposity affects adulthood earnings via education, using Finnish genotyped data and a causal mediation design. A one unit increase in childhood BMI, which corresponds to transitioning from normal weight to overweight, or from overweight to obesity, reduces adult earnings by approximately 23% through lower educational attainment. Our results further suggest that early-life adiposity has long-term economic consequences that are independent of adult weight.

BACKGROUND:Intergenerational transmission of mental disorders has been well established, but it is unclear whether exposure to a child's mental disorder increases parents' subsequent risk of mental disorders. AIMS:We examined the association of mental disorders in children with their parents' subsequent mental disorders. METHOD:In this population-based cohort study, we included all individuals with children born in Finland or Denmark in 1990-2010. Information about mental disorders was acquired from national registers. The follow-up period began when the parent's eldest child was 5 years old (for ICD-10 codes F10-F60) or 1 year old (for codes F70-F98) and ended on 31 December 2019 or when the parent received a mental disorder diagnosis, died, or emigrated from Finland or Denmark. The associations of mental disorders in children with their parents' subsequent mental disorders were examined using Cox proportional hazards models. RESULTS:The study cohort included 1 651 723 parents. In total, 248 328 women and 250 763 men had at least one child who had been diagnosed with a mental disorder. The risk of a parent receiving a mental disorder diagnosis was higher among those who had a child with a mental disorder compared with those who did not. For both parents, the hazard ratios were greatest in the first 6 months after the child's diagnosis (hazard ratio 2.04-2.54), followed by a subtle decline in the risk (after 2 years, the hazard ratio was 1.33-1.77). CONCLUSION:Mental disorders in children are associated with a greater risk of subsequent mental disorders among their parents. Additional support is needed for parents whose children have been recently diagnosed with a mental disorder.

Unemployment has been associated with negative physical health outcomes, including increased mortality risk, making it a critical issue for both policymakers and public health experts. This chapter surveys the empirical literature on the relationship between unemployment, physical health, and mortality in high-income countries. Only empirical studies conducted in high-income countries are examined to increase comparability of socioeconomic conditions, healthcare systems, and labor market dynamics. The focus is on key studies employing various empirical approaches to identify associations and causal effects, including cross-sectional correlations, panel data, and quasi-experimental designs such as plant closures and mass layoffs. The literature demonstrates a strong correlation between unemployment and various adverse physical health outcomes. However, the causal mechanisms are complex and often confounded by factors such as reverse causality and selection bias. While unemployment has been shown to exacerbate health disparities and increase mortality-particularly due to alcohol-related conditions and accidental deaths-the evidence for its direct impact on many chronic physical conditions, such as cardiovascular diseases, is less conclusive. This chapter presents the policy implications of these findings, advocating for the integration of employment and health policies to mitigate the adverse health effects of unemployment and promote social equity.

AIM:To examine the effect of the minimum legal drinking age (MLDA) on mortality and hospitalizations. DESIGN:Exploiting sharp increases in legal alcohol access at ages 18 years (beer/wine, on-premises drinking and car driving) and 20 years (off-premises spirits) in a regression discontinuity (RD) design, we estimated local linear models of all-cause and cause-specific mortality and hospitalizations in monthly data 12 months before and after each threshold, separately by sex. SETTING:Finland. PARTICIPANTS:Individuals born between 1955 and 1998 who were registered as living in Finland and observed reaching ages 18 years and 20 years in the administrative registers (n ≈ 3 000 000). MEASUREMENT:The outcomes were all-cause and cause-specific mortality and hospitalizations due to external, alcohol-related, traffic-accident and suicide-attempt causes, measured as monthly rates per 100 000 person-months. FINDINGS:Reaching the legal drinking threshold of 18 years old was associated with a substantial increase in all-cause mortality among men [RD estimate = 1.62 deaths per 100 000 person-months, 95% confidence interval (CI) = 0.86-2.38], corresponding to a 26.9% increase relative to the pre-threshold mean. This effect was driven by external causes (RD = 2.03, 95% CI = 1.25-2.82) and particularly accidental deaths (RD = 1.49, 95% CI = 0.79-2.18) and alcohol-contributed deaths (RD = 0.65, 95% CI = 0.13-1.18). Among women, the increase in all-cause mortality at age 18 years was smaller and not statistically significant (RD = 0.45, 95% CI = -0.02 to 0.92), although alcohol-contributed deaths increased from a low baseline (RD = 0.18, 95% CI = -0.02 to 0.39). At age 20 years, legal access to spirits was associated with increases in alcohol-contributed mortality among men (RD = 0.88, 95% CI = 0.19-1.56) and accidental deaths (RD = 0.54, 95% CI = 0.09-0.99), whereas among women the main effect was an increase in hospital-treated suicide attempts (RD = 1.11, 95% CI = 0.11-2.11). CONCLUSIONS:In Finland, reaching legal drinking ages appears to be associated with substantial short-run increases in mortality and morbidity, concentrated among young men and linked to external and alcohol-related causes. These findings suggest that minimum legal drinking age policies carry important public health costs that should be considered when evaluating alcohol regulation. At age 18 years, where the drinking and driving-licence thresholds coincide, both mechanisms contribute: controlling for driving-licence status reduces hospitalization estimates by approximately one quarter, but a substantial alcohol-access effect remains.

Purpose A sudden illness in family, such as cancer, is known to have significant economic effects, but the consequences of a child’s genetic disease on parental income and employment have gained much less attention. This study examines the association of having a child with neurofibromatosis 1 (NF1) and parental income and employment. Methods The Finnish cohort of 1,811 individuals with verified NF1 was cross-linked with income and employment data from Statistics Finland. The 772 parents of 390 children with sporadic NF1 and 446 parents of 229 children with familial NF1 born in 1988-2019 were compared with 7,635 and 4,465 parents of 3,862 and 2,252 matched non-NF1 individuals, respectively. Results Parents of children with NF1 experienced slower income growth and lower employment rate increases following the birth of a child compared to the parents in the comparison group. Specifically, the parents of children with NF1 had a 14-23% lower income increase and 1-3 percentage points lower rise in employment rates compared to the comparators’ parents after the child’s birth. Conclusions The effects of a child’s NF1 are gradual rather than immediate, reflecting both pre-existing labour market disparities and age-related income trends.

Background and Aims:A nutritionally balanced, age-appropriate diet is crucial for child development. The randomized Special Turku Coronary Risk Factor Intervention Project (STRIP) was a 20-year dietary intervention starting at 7 months of age. It aimed at improving fat quality and promoting healthy foods to prevent cardiovascular diseases. We examined whether the intervention was associated with university enrollment and difficulties in school-related tasks in a post hoc setting. Methods:Participants (n = 1062) were recruited at the age of 5 months between December 1, 1989, and May 30, 1992, from child health clinics in Turku, Finland, and were randomly assigned to either the intervention (n = 541) or control group (n = 521) at the age of 7 months. Children in the intervention group received personalized dietary counseling through age 20, without a fixed diet. We used linear probability models estimated by ordinary least squares (OLS) to investigate whether being in the intervention group was associated with university enrollment by age 26, and OLS regressions to examine its association with difficulties in school-related tasks at age 10. The study is reported in accordance with the CONSORT guidelines. Results:In total, 639 participants provided university enrollment data. Among males from low-education families, being in the intervention group was associated with a 32 percentage point higher likelihood of enrollment (b = 0.324, p = 0.004, 95% confidence interval (CI): 0.103, 0.546). Among all males the association was 11 percentage points (b = 0.107, p = 0.07, 95% CI: -0.007, 0.220), and among females -3 percentage points (b = -0.033, p = 0.53, 95% CI: -0.137, 0.071). The results also suggested that diet, rather than cardiovascular health, may serve as a mediator. Conclusions:Long-term dietary counseling is associated with increased educational attainment among males from low-education families. The study is registered at ClinicalTrials.gov (#NCT00223600).

Ageing populations foreground understanding how retirement affects quality of life. The inconclusive findings on retirement's quality of life impacts encourage further analyses of this association's complexity. Using waves 4 (2011) and 6 (2015) of the Survey of Health, Ageing and Retirement in Europe (N = 9249) and conditional change multilevel mixed-effects linear regressions, this study investigates how changes in three measures of social network involvement (size of close social network, frequency of contact with one's close social network, and emotional closeness with one's close social network) moderate how retirement is associated with quality of life. Our findings show that increased social network involvement concurrent with entry into retirement predicts a more beneficial retirement transition. These results are interpreted through role theory. Consequently, this study encourages societal and clinical attention to retirees' social network involvement. It also recommends further study of the nuances that affect how retirement is associated with quality of life.

Background:Clusters of self-harming behaviour among adolescents have been observed, yet population-based epidemiological evidence is lacking. This study aims to address this lack by examining the clustering of self-harming behaviour within adolescent peer networks at the population level. Methods:We used nationwide registry data on Finnish people born between January 1, 1985, and December 31, 2000, to examine whether having same-grade schoolmates who had self-harmed was associated with greater subsequent self-harm risk. Cohort members were followed up until first recorded self-harm episode, emigration, death, or December 31, 2020, whichever came first. Hazard ratios (HRs) were estimated using mixed-effects Cox proportional hazards models adjusted for a comprehensive set of individual-, parental-, school-, and area-level covariates. Findings:The cohort comprised 913,149 Finnish residents. Having same-grade schoolmates who had self-harmed between school-starting age and finishing ninth grade was associated with a higher, albeit small in magnitude, HR of subsequent self-harm over a median of 11.6 years of follow-up (HR 1.05, [95% CI 1.01-1.09]). HR was not consistently higher over follow-up time but was highest in the beginning of follow-up when the cohort members were around age 16 (1.45 [1.25-1.69]). Limiting exposure to schoolmates' self-harm episodes to 1 year consistently showed the highest risk around age 16, regardless of whether the exposure occurred in ninth grade (1.49 [1.21-1.82]) or eighth grade (1.36 [1.07-1.74]), with follow-up commencing after the respective grade. Interpretation:While we cannot rule out residual confounding, our findings suggest that self-harm may socially transmit within adolescent peer networks. The observed highest risk around age 16 suggests that external stressors associated with transitioning to new life stages at this age may moderate the impact of peer self-harm exposure. Prevention and intervention measures that consider possible peer influences on adolescents' self-harming behaviour may help reduce the public health burden of self-harm. Funding:European Research Council and Research Council of Finland.

Introduction Tobacco smoking has been associated with reduced success in the labor market, potentially due to its negative impact on labor productivity, especially in physically demanding jobs, as it affects physical fitness and performance adversely.Methods This prospective study used data from the Cardiovascular Risk in Young Finns Study survey, linked to register information on labor market outcomes and education attainment, to examine the association between tobacco smoking and long-term labor market outcomes (earnings and employment, N = 1953). Smoking levels were determined by cigarette pack-years in 2001, as reported in the survey, whereas annual earnings and employment status were tracked from 2001 to 2019.Results A one-unit increase in pack-year of smoking was associated with a 1.8% decrease in earnings (95% confidence interval [CI]: -2.6% to -0.9%) and a 0.5% reduction in years employed (95% CI: -0.6% to -0.3%). This association was pronounced among participants with lower education levels. The earnings difference was evident among younger cohorts, whereas a negative correlation with employment was observed most strongly in older cohorts among individuals with lower education.Conclusions Our findings suggest that smoking had a negative effect on earnings among the younger generation, particularly among the less well-educated. The finding of greater impacts on years of employment among the older age group, particularly among groups with low education levels, is consistent with the delayed onset of most health impacts, which may particularly affect productivity in physically demanding jobs that are more common among people with less education.Implications Adverse consequences of smoking include reduced earnings and labor market participation, particularly among less well-educated groups. Tobacco control advocates should draw attention to these consequences in arguing for effective measures to reduce smoking initiation and increase cessation in order to achieve socially optimal outcomes.

Poor information flows hamper coordination, potentially leading to suboptimal decisions in health care. We examine the effects of a nationwide policy of information integration on the quality of prescribing. We use the rollout of an electronic prescribing system in Finland and prescription-level administrative data. We find no effect on the probability of co-prescribing harmful drug combinations in urban regions. In rural regions, this probability reduces substantially, by 35 percent. The effect is driven by prescriptions from unspecialized physicians and from multiple physicians. Improving the local information environment thus enhances coordination and narrows differences in the quality of prescribing.

AIMS:There is limited research on how family-related events are associated with sustainable working life, which is defined as the absence of long-term periods of unemployment, sickness absence, or disability pension. We aimed to investigate whether family-related events, such as living with children, marriage, or divorce, are associated with sustainable working life. We also examined whether these associations could be explained by health behaviours including physical activity, smoking, and alcohol consumption, as well as by genetic factors and early life environment. METHODS:Registry data and surveys on family-related events, sustainable working life, and health behaviours of 49,611 Swedish twins in 1993-2016 were used. Cox proportional hazards regressions were applied to estimate the associations between family-related events and sustainable working life, while conditional Cox models for twin pairs accounted for confounding genetic factors and early life environment. Further, the analyses were stratified by health behaviours. RESULTS:The results showed that a change from being single living without children to being married and living with children was associated with a sustainable working life. Moreover, individuals who experienced divorce and single individuals who changed their status of living with children were less likely to experience a sustainable working life. Genetic factors and early life environment played a role in some associations. Those with healthy behaviours were more likely to have sustainable working lives despite changes in family life, compared with those with unhealthy behaviours. CONCLUSIONS:Family formation (i.e. change from being single to getting married and having children) is associated with sustainable working life, whereas divorce seems to decrease the possibility of sustainable working life. Health behaviours, as well as genetic factors and early life environment, also need to be considered with changes in family life to support sustainable working life.

PurposePrevious studies document the clustering of major psychiatric disorders (MPDs) - schizophrenia, bipolar disorder, depression, and anxiety - among siblings. Few studies have, however, examined whether MPDs during childhood and early adulthood are associated with siblings' future socioeconomic status (SES).MethodsThis cohort study included 57,537 full siblings, 4653 paternal, and 5053 maternal half-siblings of individuals with MPDs (affected probands) born in Finland between 1970 and 1990. We defined the reference groups as identical types of siblings of individuals without an MPD diagnosis (unaffected probands) and followed both siblings of the affected and unaffected probands until December 31, 2020. MPDs diagnosed among the affected probands at ages 5-25 was obtained from the Finnish Care Register. Their siblings' SES was measured based on employment status, annual disposable income, and educational achievement. Logistic regression, median regression, and generalized estimating equations (GEE) were used to estimate the associations.ResultsCompared to the siblings of the unaffected probands, the odds of unemployment at the end of follow-up were 35% higher (95% CI: 1.31-1.39) in full siblings of affected probands with an MPD. Full siblings of affected probands were also more likely not to achieve a higher education level (aOR: 1.28, 95% CI 1.24-1.31). The median annual disposable income was 1255.9 EUR lower (95% CI: -1385.6, -1126.3) in full siblings of affected probands. Similar but weaker associations were observed in maternal and paternal half-siblings. Results from GEE models using repeated measurements of income and unemployment were similar.ConclusionOur findings suggest that the socioeconomic consequences associated with MPDs extend to siblings.

Background: With tightening healthcare budgets, assessing the economic impact of mental health treatments is crucial for informing policy measures. Sociodemographic factors, such as gender, are associated with these outcomes. We studied the gender-specific impact of long-term rehabilitative psychotherapy on the income of working-age individuals who had experienced sick leave due to common mental disorders (CMD). Methods: We analyzed register data from 2010 to 2019, covering a 33 % random sample of the Finnish workingage population (aged 18-55 in 2010). The sample included 32,558 individuals with their first compensated sickness absence for CMD during 2011-2015. Of these, 4592 (76.1 % women) began psychotherapy, the rest formed the control group. We used difference-in-differences regression with inverse probability weighting to estimate the average treatment effects on the treated (ATT) for women and men relative to the timing of treatment. Results: The average effect was negligible for women (ATT: 0.01, 95 % Confidence Interval: -0.03 - 0.06), while men experienced a substantial 27 % increase in income (ATT: 0.27, 95 % CI: 0.13-0.42). Event-study estimates indicated that the effect on men's income was immediate and long-lasting. Limitations: Analysis of our findings is restricted to observed register-based covariates including work disability and employment status, and individual- and time-fixed effects. Conclusions: Among those who had received compensated sick leave due to CMD, men experienced an immediate and long-lasting increase in income after undergoing psychotherapy, while no such effect was observed for women. Further research is needed to explore factors such as selection for treatment, occupational segregation, and treatment engagement among recipients.

We examine the heterogeneous effects of 14-21 euro copayments on primary care general practitioners (GPs) visits in Finland. Our study focuses on a triage-based appointment system in public primary care. Using an age-based regression discontinuity (RD) design and leveraging variation across Finnish municipalities in whether the copayment is charged, we analyze the effects at the 18th birthday, when previously exempted adolescents become subject to copayments. Using nationwide administrative data from 2011 to 2019, we find that GP visits decrease in the copayment municipalities by 4-5 %. The reductions are largest for the bottom 20 % of the equivalized family disposable income distribution: their GP use decreases by 0.08-0.10 annualized visits (7-10 %). Unexpectedly, the effects are also larger than average (albeit rather temporary) for the top 50 %, showing reductions of 6-8 %. Compared to earlier studies focusing on moderate copayments and different populations, our effect estimates are smaller, and the heterogeneity by income level is weaker.

We propose a method of studying the value of insurance. For this purpose, we analyze well-being of the same individuals, comparing sick and healthy years in German panel survey data on life satisfaction. To impose structure on the income–well-being gradient, we fit a flexible utility function to the data, focusing on the differences in marginal utility in the sick and the healthy state, by allowing for a “fixed cost of sickness”. We find that marginal utility of income is higher in the sick state. We use our estimates to gauge the value of sickness insurance for Baily-Chetty type optimal policy calculations. We also show that the income–well-being gradient has steepened over time in Germany and we use the fitted model to characterize this change.

Emerging strands of research have examined the family spillover effects of health shocks, usually focusing on labour market outcomes. However, the results have been inconclusive and there is only little evidence on the longer term consequences of health shocks or the mechanisms behind the spillover effects. We analyse the short- and long-term effects of cancer on the healthy spouse's labour supply and mental health by gender and relative income status within the couple (i.e., the breadwinner type). We use full population register data on all cancer patients and their cohabiting partners in Finland over the period 1995-2019. Our identification strategy is based on the quasi-random variation in the timing of the cancer diagnosis and a dynamic difference-in-differences approach. We find two main results. First, cancer increases female spouses' employment. This result is consistent with the added worker effect, although we find the magnitude of the increase in annual earnings to be negligible. By contrast, among male spouses, earnings decrease as a consequence of a spouse's cancer. Second, among women, there is heterogeneity in the effects in terms of the breadwinner status, which is especially notable in the long-term. The results show that the added worker effect is visible only among secondary earners and the effect seems to hold only when the cancer patient dies. Secondary earner women also suffer more from psychiatric symptoms during bereavement. Consequently, we argue that the breadwinner status before the health shock is a neglected factor influencing the effects of health shocks in families, and that family-level specialisation between spouses alters substantially over time in response to a health shock.

Spillover effects of mental disorders on family members are well-documented, but their impact on adolescent peers remains unclear. We investigated whether having classmates with a diagnosed mental disorder in the ninth grade of lower secondary school (ages 15-16) was associated with greater likelihood of dropping out of upper secondary education (ages 16-19). We used registry data on Finnish people born between Jan 1, 1985, and Dec 31, 1997. After lower secondary school, 378,453 cohort members started general (academic) upper secondary school and 284,713 started vocational upper secondary school. Using causal mediation analysis, we disentangled the total effect of having ninth-grade classmates with a diagnosed mental disorder on upper secondary dropout into direct spillover effect and indirect effect mediated by the individual's own mental disorder diagnosis received during upper secondary education. For academic students, having one ninth-grade classmate diagnosed with a mental disorder was associated with 6% greater likelihood of dropping out of upper secondary education (95% CI 3-9%); two diagnosed classmates, with 10% greater likelihood (6-14%); and three or more diagnosed classmates, with 16% greater likelihood (10-22%). For vocational students, the respective increases in likelihood were 5% (2-8%), 11% (6-15%), and 24% (18-32%). For both academic and vocational students, the indirect effect of dropping out mediated by one's own mental disorder diagnosis was notably smaller than the direct effect attributable to having diagnosed ninth-grade classmates. Our results suggest that mental disorders may have spillover effects on the educational attainment of socially connected adolescent peers.