Sensitivity of familial high-risk approaches for the identification of future bipolar disorder: A longitudinal total birth cohort study

BACKGROUND: Extensive research has demonstrated an increased risk of bipolar disorder (BD) in individuals with familial high-risk for BD (FHR-BD). To date, however, it has not been demonstrated what proportion of all BD cases are captured by the FHR-BD approach. This is essential information because it identifies the upper limit of BD cases that could be prevented using the FHR-BD approach, if we had an effective preventive intervention. METHODS: All individuals born in Finland 1987-1992 (index children) were identified in health care registers and linked with maternal and paternal data. Of all BD diagnoses in the population by age 29, we assessed the proportion that occurred in individuals with a parental history of BD (FHR-BD). As an additional novel transdiagnostic approach, we also assessed the proportion of BD diagnoses that occurred in individuals who had a parental history of psychiatric admission for any reason (transdiagnostic familial risk; TDFR-BD). RESULTS: Of all index children (n = 368,779), 2.4 % (n = 8673) met FHR-BD criteria and 15.3 % (n = 56,483) met TDFR-BD criteria. Of all individuals diagnosed with BD, 7.8 % (CI: 6.9-8.7, n = 307) met FHR-BD criteria and 27.2 % (CI: 25.8-28.6 %, n = 1061) met TDFR-BD criteria prior to diagnosis. Absolute risk of BD in FHR-BD children and TDFR-BD children was 4.1 % (CI: 3.6-4.6 %, n = 307) and 2.2 % (CI: 2.0-2.4 %, n = 1061), respectively. CONCLUSIONS: Only a small proportion (7.8 %) of BD cases are identified using the FHR-BD approach. Our findings highlight the importance of prediction research investigating not only absolute or relative risk, but also the sensitivity of different approaches to identifying risk.