Previous studies reported an association between advanced paternal age at birth and increased risk for schizophrenia and bipolar disorder. While some hypothesize that this association is caused by de-novo mutations in paternal spermatozoa, others cite factors associated with psycho-social characteristics of fathers who have children at a late age. This study aims to test these hypotheses.

A historical-prospective, population-based cohort study, performed by linking the Israeli Draft Board Registry and the Israeli National Psychiatric Hospitalization Registry (N = 916 439; 4488 with schizophrenia, 883 with bipolar disorder). Odds ratios (OR) and two-sided 95% confidence intervals (CI) were calculated by logistic regression models, using paternal age as predictor and risk for later hospitalizations for schizophrenia or bipolar disorder as outcome measure. Models were first fitted unadjusted, then adjusted for paternal age at birth of the first child.

In the unadjusted model, offspring of fathers aged 45 and above at birth had increased risk of schizophrenia (OR = 1.71, 95% CI 1.49–1.99) and bipolar disorder (OR = 1.63, 95% CI 1.16–2.24). However, taking into account paternal age at birth of first child, advanced paternal age was no longer associated with increased risk of schizophrenia (OR = 0.60, 95% CI 0.48–0.79) or bipolar disorder (OR = 1.03, 95% CI 0.56–1.90).

Controlling for paternal age at birth of the first offspring, advanced paternal age does not predict increased risk for schizophrenia or bipolar disorder. These data indicate that the association between advanced paternal age and having an offspring with schizophrenia and bipolar disorder is likely due to psychos-social factors, or common genetic variation associated with delayed initial fatherhood.

Advanced paternal age at birth has been linked to several psychiatric disorders in offspring (e.g. schizophrenia) and genetic mechanisms are thought to underlie these associations. This study is the first to investigate whether advanced paternal age at birth is associated with eating disorder risk using a twin study design capable of examining both phenotypic and genetic associations.

In a large, population-based sample of female twins aged 8–17 years in mid-puberty or beyond (n = 1722), we investigated whether advanced paternal age was positively associated with disordered eating symptoms and an eating disorder history [i.e. anorexia nervosa (AN), bulimia nervosa (BN) or binge eating disorder (BED)] in offspring. Biometric twin models examined whether genetic and/or environmental factors underlie paternal age effects for disordered eating symptoms.

Advanced paternal age was positively associated with disordered eating symptoms and an eating disorder history, where the highest level of pathology was observed in offspring born to fathers ⩾40 years old. The results were not accounted for by maternal age at birth, body mass index (BMI), socio-economic status (SES), fertility treatment or parental psychiatric history. Twin models indicated decreased genetic, and increased environmental, effects on disordered eating with advanced paternal age.

Advanced paternal age increased risk for the full spectrum of eating pathology, independent of several important covariates. However, contrary to leading hypotheses, environmental rather than genetic factors accounted for paternal age–disordered eating associations. These data highlight the need to explore novel (potentially environmental) mechanisms underlying the effects of advanced paternal age on offspring eating disorder risk.