BackgroundEarly retirement caused by disability pension (DP) due to common mental disorders (CMDs) is frequent in European countries. Inadequate treatment, e.g., suboptimal antidepressant (AD) medication before DP can be crucial in such DP. This explorative study aimed to disentangle trajectories of AD based on defined daily dose (DDD) before and after granted DP, and to characterize the trajectories by socio-demographics and medical factors.MethodsAll 4642 individuals in Sweden aged 19–64 with incident DP due to CMD in 2009–2010 were included. Trajectories of annual DDDs of AD were analysed over a 6-year period by a group-based trajectory method. Associations between socio-demographic or medical factors and different trajectories were estimated by chi2-test and multinomial logistic regression.ResultsFive trajectories of ADs were identified. Three groups, comprising 34%, 34%, and 21% of the cohort, had constant AD levels before and after DP with mean annual DDDs of 29, 234, and 580, respectively. Two groups, each including 6% of the cohort, had increasing levels of DDDs, levelling off at around 1150 and 785 DDDs after DP. Particularly age, outpatient care due to mental diagnoses and DP diagnoses were significantly associated with different trajectories (p < 0.05). All the groups had a larger proportion of older individuals (> 50%, 45–64 years), except for the ‘increasing low’ group, where younger individuals were in majority (> 60%, 18–44 years), who more frequently exited labour market due to ‘anxiety disorders’, with lower education and more specialised healthcare before DP than the other groups.ConclusionThe heterogeneity among the five trajectory groups was partly explained by age, the severity of the mental disorder and the DP diagnoses. DDDs of ADs, though on different levels, varied marginally before and after granted DP in the majority. Moreover, AD levels were very low in one third of the individuals. Early identification and focus on the ‘increasing low’ group might be important in order to identify individuals at risk for further increase in annual DDDs of ADs even after granted DP, and might also contribute in prevention of DP. Further detailed research regarding different groups is warranted.