Background Notwithstanding difficulties in ascertaining intent, aetiological studies on adult poisoning typically focus on intentional or unintentional events as distinct entities. This study investigated the predictors of hospital admissions or deaths for intentional and unintentional poisoning in adult New Zealanders.

Methods The 22,389 participants aged 16 and older comprised the New Zealand Blood Donors Health Study (NZBDHS) cohort who completed a baseline self-report questionnaire (including demographic, personal health, psychosocial and lifestyle information) at recruitment in 1998/1999. Outcome data on poisoning-related admissions and deaths up to 31 December 2014 were collected prospectively through electronic record linkage of participants’ unique identifiers to national mortality and morbidity databases. Baseline characteristics associated with intentional and unintentional poisoning at follow-up were investigated using Cox proportional hazards analysis.

Results During the median follow-up period of 16.8 years (359,018 person-years), 437 poisoning events (315 intentional, 124 unintentional) were identified in 310 cases. Multivariable models revealed that both intentional and unintentional poisoning at follow-up were associated with depressive symptoms (intentional poisoning: adjusted HR = 2.05, 95% CI: 1.20–3.52; unintentional poisoning: adjusted HR = 1.58, 95% CI: 1.01–2.49) and suicidal ideation at recruitment (intentional poisoning: adjusted HR = 5.76, 95% CI: 3.32–9.97; unintentional poisoning: adjusted HR = 2.45, 95% CI: 1.53–3.94). Illegal drug use was a risk factor for future unintentional poisoning events.

Conclusions The findings of this large prospective injury cohort study suggest that interventions addressing mental health problems have the potential of reducing serious poisoning events irrespective of intent. Whether the findings also represent misclassification of intentional injuries as unintentional events is unclear.