Background: Among older people intentional poisoning outnumber unintentional ones. While there are indications that time trends differ by poisoning intent, studies are scarce. We assessed how the annual prevalence of intentional and unintentional poisoning changed over time, overall and by demographic groups.

Methods: We conducted a national open cohort study of individuals aged 50-100 years, resident in Sweden during 2005-2016. Individuals were followed up in population-based registers for their demographic and health attributes from 2006-2016. Annual prevalence of hospitalization and death by poisoning intent (unintentional vs. intentional or undetermined; ICD-10 definitions) were compiled for the categories of four demographic attributes (age, sex, marital status, and birth cohort "baby boomers"). The time trends were assessed by multinomial logistic regression with year as an independent variable.

Results: The annual overall prevalence of hospitalization and death by intentional poisonings consistently exceeded that of unintentional poisonings. There was a significant downward trend in intentional poisonings but not in unintentional ones. This difference in trends also applied when considering men and women separately, married and unmarried people, the young-old individuals (but not the older- or oldest-old ones), and the baby boomers and non-baby boomers. The largest demographic differences within intent were found between married and unmarried people, and the smallest one between men and women.

Conclusion: As expected, the annual prevalence of intentional poisonings considerably exceed that of unintentional ones among Swedish older people. The recent trends reveal a significant reduction of intentional poisonings, consistent across a range of demographic attributes. The scope for action regarding this preventable cause of mortality and morbidity remains considerable.