Importance: Cerebral palsy (CP) is considered a paediatric condition despite most people living into adulthood. Due to this we lack evidence in adults with CP, this includes a paucity of research examining mental health in this population.

Objectives: Determine the risk of depression and anxiety in adults with a diagnosis of CP compared with an age-, sex-, and practice-matched reference group of adults without CP, using primary care data.

Design, setting and participants: Retrospective longitudinal cohort study set in UK primary care. Data were analysed using Cox proportional hazards regression analyses adjusted for chronic conditions and visits to their physician. The study period ran from 1987 to 2015. Data for 1,705 adults aged 18 or older with CP and 5,115 matched adults who did not have CP were extracted. CP was identified using diagnostic codes, and each person with CP was compared with 3 age, sex and practice matched controls.

Exposure: Diagnosis of CP, with a second analysis accounting for co-morbidity of intellectual disability (ID).

Main outcomes: Time to diagnosis for depression or anxiety following the date of entry into the study in adults with CP (with and without ID) when compared with matched controls.

Results: The mean age of the CP and matched group was 33.3 and 46.8% (n=798) were female. People with CP had an increased adjusted hazard of depression (HR 1.28, 95% CI: 1.09-1.51) and anxiety (HR 1.40, 95% CI: 1.21-1.63) when compared with the matched reference group. When we accounted for ID co-morbidity there were 363 adults with CP who also had ID (mean age 32.1, 47.6% (n=159) female) and 1342 adults with CP who did not have ID (mean age 33.6, 43.8% (n=639) female). Only those people with CP and no co-morbid ID had a higher risk of incident depression (HR 1.44: 95% CI 1.20-1.72) and anxiety (HR 1.55: 95% CI 1.28-1.87) than their matched controls.

Conclusions: The results demonstrate that adults with CP have an increased risk of depression or anxiety. In particular, our results indicate that this association is driven largely by those individuals with CP with no co-occurring ID. Future work is needed in community-based samples in order to fully elucidate the causal mechanisms driving these associations.