• bipolar disorder;
  • epidemiology;
  • parental age;
  • U-shaped


Studies on the association between parental age and bipolar disorder (BPD) are scarce and with inconsistent findings. The aim of this study was to examine the association of parental age and age difference between parents with risk of BPD in offspring.


This nested case–control study identified 1,861 cases of individuals with BPD born in Finland during 1983–1998 and diagnosed by the end of 2008, and 3,643 sex- and date of birth-matched controls from nationwide population-based registers. Conditional logistic regression was used to examine the association adjusting for potential confounding due to age of the other parent, parental psychiatric history, educational level, and place of birth.


A U-shaped association of unadjusted odds ratios (ORs) for BPD risk was seen in different paternal age categories, with the odds increasing at both ends of the age spectrum. In the adjusted analyses, offspring of fathers aged ≥50 years had a 2.8-fold increased odds [OR = 2.84, 95% confidence interval (CI): 1.32–6.12] of BPD as compared to those with fathers aged 30–34 years. The odds were increased 1.3-fold (OR = 1.35, CI: 1.06–1.72) in fathers aged 20–24 years. No significant association was found between maternal age and BPD in the adjusted analyses. Age difference between parents was not associated with BPD.


The increased risk of BPD in offspring of the youngest and oldest fathers in the study suggests the involvement of different biological and psychosocial factors at the two ends of the paternal age spectrum. These findings may be significant in the context of advancing parental age in recent times.