Fertility trends of women with serious mental illness in the United Kingdom 1992–2017: A primary care cohort study using the clinical practice research datalink

•The fertility trends of women with affective and non-affective psychoses are examined for the first time.•The live-birth rate ratio of women with psychoses has not changed (1992–2017).•The pregnancy rate ratio increased for women with affective psychoses, only.•There is no fertility benefit associated with second-generation antipsychotics.•Pregnancy rates double immediately after discontinuing any antipsychotic. Changes in care may mean women with serious mental illness (SMI) are more fertile. We investigated 1) the live-birth and pregnancy rate of women with and without SMI over time, 2) the likelihood of pregnancy when using second or first-generation antipsychotics. Retrospective cohort study of women (15–45 years) registered in Clinical Practice Research Datalink (CPRD) general practices between 1992 and 2017. Each analysis year, women with SMI (affective and non-affective psychotic disorder) were matched with up to four women with no record of SMI on age, calendar year and general practice. Pregnancy and live-birth rates and the rate ratio (RR) comparing women with and without SMI were estimated. The stability of the RR between years was tested. For women with SMI, the pregnancy rates when on or off first or second-generation antipsychotics were calculated and compared using Poisson regression models. In total, 12,524 women with SMI were matched to 50,074 women without SMI, median age 34 [IQR 28–39] years. Between 1992 and 2017 women with SMI had 50% fewer live-births than women without SMI (RR 0..50, 95%CI 0.45–0.55). The pregnancy rate ratio increased from 0.64 (95%CI 0.48–0.86) (1992–1994) to 1.00 (95%CI 0.81–1.22) (2016–2017), (p