Mental health and behavioural problems in adolescents conceived after ART

Document Type

Journal Article

Publication Title

Human reproduction





First Page


Last Page


PubMed ID



Oxford Academic


School of Medical and Health Sciences


National Health and Medical Research Council / Australian Government Research Training Program Scholarship

Grant Number

NHMRC Number : 1042269


Wijs, L. A., Doherty, D. A., Keelan, J. A., Burton, P., Yovich, J. L., Robinson, M., & Hart, R. J. (2022). Mental health and behavioural problems in adolescents conceived after ART. Human Reproduction, 37(12), 2831-2844. https://doi.org/10.1093/humrep/deac214


Study question: Does mental health and behaviour differ between those conceived with and those conceived without ART? Summary answer: Our study observed less externalizing behaviour (delinquent/aggressive), and more parent-reported internalizing behaviour, as well as more (clinical) depression at age 14 years, in adolescents conceived after ART compared to their non-ART counterparts. What is known already: Health outcomes of ART-conceived offspring may differ from those conceived without ART, and previous studies have reported differences in behaviour and mental health, particularly in childhood. Study design, size, duration: The Growing Up Healthy Study (GUHS) is a prospective cohort study, investigating the long-term health of offspring conceived after ART (aged 14, 17 and 20 years), in the two operational fertility clinics in Western Australia 1991-2001 (n = 303). Their long-term health outcomes were compared to those of offspring conceived without ART from the Raine Study Generation 2 (Gen2) born 1989-1991 (n = 2868). Both cohorts are representative of the local adolescent population. Participants/materials, setting, methods: Mental health parameters and behaviour were assessed at ages 14 and 17 years, through the parent completed 'Child Behaviour Checklist' (CBCL; ART versus non-ART: age 14 years: N = 150 versus N = 1781, age 17 years: N = 160 versus N = 1351), and the adolescent completed equivalent 'Youth Self-Report' (YSR; age 14 years: by N = 151 versus N = 1557, age 17 years: N = 161 and N = 1232). Both tools generate a T-score (standardized for age and sex) for internalizing (withdrawn, somatic complaints, anxious/depressed), externalizing (delinquent/aggressive behaviour) and total behaviour. Adolescents also completed the 'Beck Depression Inventory for Youth' (BDI-Y; age 14 years: N = 151 versus N = 1563, age 17 years: N = 161 versus N = 1219). Higher scores indicate poorer mental health and behaviour on all the above tools. Parent-reported doctor-diagnosed conditions (anxiety, behavioural problems, attention problems and depression) were also univariately compared between the cohorts. In addition, univariate comparisons were conducted between the GUHS adolescents and Gen2 adolescents born to subfertile parents (time to pregnancy > 12 months), as well as between offspring born to subfertile versus fertile parents within the Gen2 cohort. A subgroup analysis excluding offspring born preterm ( < 37 weeks' gestation) or at low birthweight ( < 2500 g) was also performed. Generalized estimating equations that account for correlated familial data were adjusted for the following covariates: non-singleton, primiparity, primary caregiver smoking, family financial problems, socio-economic status and both maternal and paternal ages at conception. Main results and the role of chance: At both 14 and 17 years of age, ART versus non-ART-conceived adolescents reported lower mean T-scores for externalizing problems (age 14 years: 49 versus 51, P = 0.045, age 17 years: 49 versus 52, P < 0.001). A similar effect was reported by parents, although not significant (age 14 years: P = 0.293, age 17 years: P = 0.148). Fewer ART-conceived adolescents reported a T-score above the clinical cut-off for externalizing behaviour ( ≥ 60; age 14 years: 7.3 % versus 16.3 %, P = 0.003, age 17 years: 8.1 % versus 19.7 %, P < 0.001). At both ages, no differences in internalizing behaviour were reported by adolescents (age 14 years: P = 0.218, age 17 years: P = 0.717); however, higher mean scores were reported by parents of the ART-conceived adolescents than by parents of the non-ART conceived adolescents (age 14 years: 51 versus 48, P = 0.027, age 17 years: 50 versus 46, P < 0.001). No differences in internalizing behaviour above the clinical cut-off (T-score ≥ 60) were observed. At age 17 years, parents who conceived through ART reported higher total behaviour scores than those parents who conceived without ART (48 versus 45, P = 0.002). At age 14 years, ART versus non-ART-conceived adolescents reported significantly higher mean scores on the BDI-Y (9 versus 6, P = 0.005); a higher percentage of adolescents with a score indicating clinical depression ( ≥ 17; 12.6 % versus 8.5 %, aOR 2.37 (1.18-4.77), P = 0.016), as well as more moderate/severe depression ( ≥ 21; 9.3 % versus 4.0 %, P = 0.009). At age 17 years, no differences were reported on the BDI-Y. There was also a higher percentage of parent-reported doctor-diagnosed anxiety in the ART cohort (age 14 years: 8.6 % versus 3.5 %, P = 0.002, at age 17 years: 12.0 % versus 4.5 %, P < 0.001). Removing adolescents born preterm or at low birthweight did not alter the above results. Comparing outcomes between GUHS adolescents and Gen2 adolescents born to subfertile parents, as well as between those born to subfertile versus fertile parents within Gen2, did not alter results for CBCL and YSR outcomes. Those born to subfertile parents showed higher rates of clinical depression than those born to fertile parents at age 14 years (13.7 % versus 6.9 %, P = 0.035). Limitations, reasons for caution: The main limitation of the study is the time difference between the GUHS and Gen2 assessments. Even though we have adjusted for covariates, additional socio-economic and lifestyle factors affecting behaviour and mental well-being could have changed. We were unable to differentiate between different types of ART (e.g. IVF versus ICSI), owing to the low number of ICSI cycles at the time of study. Fertility sub-analyses need to be replicated in larger cohorts to increase power, potentially using siblingship designs. Lastly, selection bias may be present. Wider implications of the findings: The reported lower prevalence of externalizing behaviour (delinquent/aggressive), and higher prevalence of internalizing behaviour, as well as more (clinical) depression at age 14 years, in ART versus non-ART-conceived adolescents, is in line with some previous studies, mostly conducted in childhood. It is reassuring that differences in the rates of depression were not observed at age 17 years, however, these findings require replication. As the use of ART is common, and mental health disorders are increasing, knowledge about a potential association is important for parents and healthcare providers alike.



Access Rights

subscription content