Antidepressant use and risk of adverse outcomes in people aged 20 to 64

Talk Code: 
Carol Coupland
Trevor Hill, Richard Morriss, Michael Moore, Antony Arthur, Julia Hippisley-Cox
Author institutions: 
University of Nottingham, University of Southampton, University of East Anglia


Antidepressants are one of the most commonly prescribed medications in primary care. More than 57 million antidepressant prescriptions were issued in England in 2014, compared with 29 million 10 years previously. There are a large number of different antidepressants available, including various types of selective serotonin reuptake inhibitor (SSRI), tricyclic and related antidepressants and several other antidepressant drugs. National guidelines propose that since these antidepressants have broadly equal efficacy the choice of an antidepressant should mainly depend on consideration of adverse effects, but that when an antidepressant is prescribed, it should normally be an SSRI. Evidence of adverse effects of different antidepressants however remains limited, and is often from short term trials in selected patients. We therefore carried out a cohort study to assess a range of adverse outcomes which could be associated with antidepressant treatment.


A cohort of 238,963 patients aged 20 to 64 with a first recorded diagnosis of depression between 1/1/2000 and 31/07/2011 was identified using the QResearch primary care database and followed up to 31/07/2012. Information was extracted on prescriptions for antidepressants during follow-up, and on potential confounding variables. Cox proportional hazards models were used to estimate adjusted hazard ratios for the associations between falls, fractures, gastro-intestinal bleeds and adverse drug reactions over five years follow-up, including analyses of antidepressant class, dose, individual drug type and duration of use.


There were significant differences between the antidepressant drug classes for fracture and adverse drug reactions. The fracture rate was significantly lower for tricyclic and related antidepressants than for selective serotonin reuptake inhibitors (adjusted hazard ratio 0.71, 95% CI 0.61 to 0.82) but rates of adverse drug reaction were significantly higher for tricyclic and related antidepressants and other antidepressants than for selective serotonin reuptake inhibitors (adjusted hazard ratios 1.54, 95% CI 1.25 to 1.88 and 1.61, 95% CI 1.22 to 2.12 respectively). For individual antidepressant drugs there were significantly lower fracture rates for amitriptyline and dosulepin compared with citalopram, but higher rates of adverse drug reactions for amitriptyline, lofepramine and venlafaxine. Trazodone was associated with a significantly higher rate of upper gastro-intestinal bleeding than citalopram (adjusted hazard ratio 2.73, 95% CI 1.44 to 5.17), .


This study has found some differences in rates of serious adverse outcomes between different antidepressant classes and individual drugs in people aged 20 to 64 with a diagnosis of depression. The results are susceptible to residual confounding and indication bias but indicate potential associations which should be considered with individual patients when antidepressants are prescribed.

Submitted by: 
Carol Coupland
Funding acknowledgement: 
This project was funded by the National Institute for Health Research (NIHR) School for Primary Care Research (project No 81). This abstract presents independent research funded by the National Institute for Health Research (NIHR). The views expressed are those of the authors and not necessarily those of the NHS, the NIHR or the Department of Health.