Skip Navigation


Age and Ageing Advance Access originally published online on May 30, 2007
Age and Ageing 2007 36(4):449-454; doi:10.1093/ageing/afm056
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
36/4/449    most recent
afm056v1
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Hansen, D. G.
Right arrow Articles by Vach, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hansen, D. G.
Right arrow Articles by Vach, W.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

Copyright © The Author 2007. Published by Oxford University Press on behalf of the British Geriatrics Society.

Increased use of antidepressants at the end of life: population-based study among people aged 65 years and above

Dorte Gilså Hansen1, Jens-Ulrik Rosholm2, Anthony Gichangi3 and Werner Vach4

1 Research Unit for General Practice, University of Southern Denmark, JB Winsløws Vej 9A, DK-5000 Odense C, Denmark
2 Department of Geriatrics, Odense University Hospital, Sdr. Boulevard 29, DK-5000 Odense C, Denmark
3 Research Unit for General Practice, University of Southern Denmark, JB Winsløws Vej 9A, DK-5000 Odense C, Denmark
4 Department of Statistics, University of Southern Denmark, JB Winsløws Vej 9B, DK-5000 Odense C, Denmark

Address correspondence to: Dorte Gilså Hansen. Email: dgilsaa{at}health.sdu.dk


   Abstract
Top
 Abstract
Introduction
 Methods
 Results
 Discussion
 key points
 Conflict of interest
 Notes
 References
 
Background: the new antidepressants are generally effective and safe for older people, but may have serious side-effects. The use has been rapidly increasing, but focus on upper age groups has been limited. The pattern of antidepressant use as death approaches has never been analysed.

Objective: to analyse the use of antidepressants among individuals aged 65 years and above with respect to time trends, age and proximity to death.

Design: population-based prescription study.

Setting: the County of Funen, Denmark, 1992–2004 (~470.000 inhabitants).

Results: the 1-year prevalence of antidepressants increases steadily over time in all age groups. Among the 65+ year-olds it also increases with age and differs substantially between the youngest and the oldest. Very high prevalences are observed: 26.8% among females 85–89 years old and 17.5% among males 85 years and above in 2004. In all age groups the use of antidepressants increases substantially with proximity to death in the last 3 years of life. In the last phase of life the use is independent of whether the patient dies at age 65 or 90 about 33% of females and 25% of males receive antidepressants in the last 6 months.

Conclusions: the use of antidepressants among 65+ year-olds increases with age and proximity to death to very high levels. Future studies may clarify the problems and diagnoses giving rise to the use of antidepressants for such a large proportion of older people, and especially the problems giving rise to treatment as death approaches.

Keywords: antidepressive agents, drug prescriptions, aged, death, pharmacoepidemiology, elderly


   Introduction
Top
 Abstract
 Introduction
Methods
 Results
 Discussion
 key points
 Conflict of interest
 Notes
 References
 
The consumption of antidepressants has increased dramatically over the years. This is particularly due to the selective serotonin reuptake inhibitors (SSRIs) introduced in the late 1980s and a few newer agents/substances. The indications for these drugs have gradually changed, and in addition to major depression, include mild depression, neuropathic pain, panic disorders, most anxiety disorders, eating disorders and treatment of agitation in patients with dementia. Major depression and depressive symptoms are, however, suggested to be the major indication [1]. The SSRIs are considered safer than the tricyclic antidepressants. Side-effects particularly serious for older people have, however, been described, e.g. hyponatraemia [2], upper gastrointestinal bleeding [3], and falls [4]. The high prescribing rates have been of great clinical and public health importance and an intensive debate on the justification of the increased use of antidepressants has taken place.

Evidence from psychiatric morbidity surveys suggests that the current level of major depression in the general population is about 5%, but higher in older age groups. European studies have reported point prevalences of 15–20% for people aged 80+ years [5]. For all age groups, many depressed people go undiagnosed and among those diagnosed antidepressants are often not initiated [5–7]. The use of antidepressants for treatment of depression is thus expected to be lower than the prevalence of depression. Drug utilisation studies based on antidepressant prescription data have been carried out in many countries including Denmark [8, 9]. Details about prescribing levels for older age groups have, however, been limited. Upper age groups have been very wide (e.g. 65+ [10], 75+ [11]) and the number of persons very small.

Patient characteristics including comorbidity and prognosis affect prescribing decisions, even for conditions unrelated to a patient's main medical problem [12]. Prescribing of antidepressants may change as life expectancy shortens. Specific prescribing patterns at the end of life may influence prevalence figures and particularly those for older age groups with high mortality. No previous study has analysed the use of antidepressants in relation to proximity to death despite the fact that it may bring important knowledge to our understanding of the use of these drugs among older people.

Based on pharmacy data covering a population of ~470,000 people during 1992–2004, the aim of this paper was to analyse the use of antidepressants among individuals aged 65 years and above with respect to time trends, age, and proximity to death.


   Methods
Top
 Abstract
 Introduction
 Methods
Results
 Discussion
 key points
 Conflict of interest
 Notes
 References
 
The study was based on prescription data and demographic information about all inhabitants in the County of Funen (~470, 000). All data were retrieved from Odense University Pharmacoepidemiologic Database (OPED).

Data
Demographic data containing gender, birth, date of migration and death for all persons living in the county were retrieved for 1992–2004 and linked to individual records on all antidepressants redeemed at all pharmacies in the county. Pharmacy records included date of redemption, patient identification, date of birth, gender, and drug (ATC-code1). Of the 1,420,934 prescriptions, the 4.2% redeemed by persons living outside the county at the time of redemption were excluded. The antidepressants were classified into: (i) Tricyclic antidepressants (TCAs), (ii) SSRIs, and (iii) other antidepressants. No information on indication or dosing was available.

Statistical analysis
The 1-year prevalence of antidepressant drug use was defined as the number of subjects with at least one prescription during a calendar year per 100 inhabitants. Age-specific prevalences were defined with respect to the population at 1 January each year. Due to high mortality in older age groups many subjects do not survive a calendar year. Naive prevalence estimates counting the number of subjects with at least one prescription during a calendar year tend to underestimate the prevalence among age groups with high mortality and comparison across age groups becomes questionable. To overcome this problem, we calculated the prevalence of prescribing as 1 minus the probability of remaining drug free during the year by use of the Kaplan–Meier method. Death and migration were handled as censoring events.

To analyse if the observed age relationship in the general use of antidepressants could be attributable to specific prescribing patterns close to death, we repeated the computations of the 1-year prevalence excluding individuals 1 and 2 years prior to death.

To analyse the effect of proximity to death, we used a retrospective cohort design including people who died in 2001–04. We constructed six observational 6-month windows backwards in time, beginning with patients' dates of death and ending 3 years prior to death. The proportion of users in each interval was computed. This analysis was restricted to people living in the county during the last 3 years of life. Finally, it was repeated for people who died 1995–98.

All analyses were conducted for males and females separately using age group classification of 20–64, 65–69, ..., 90–94, and above 95 years. The age group below 65 years was included for comparison. Due to small numbers some upper age groups were pooled for specific analyses. A 95% confidence level was used in constructing confidence intervals (CI).


   Results
Top
 Abstract
 Introduction
 Methods
 Results
Discussion
 key points
 Conflict of interest
 Notes
 References
 
As illustrated in Figure 1, the 1-year prevalence of antidepressant use increased steadily from 1992 to 2004 in all age groups. A clear age relationship was observed, and the 1-year prevalence peaked at 26.8% among females 85–89 years old and 17.5% among males 85 years and above. The absolute increase was most pronounced among females, and among older compared with younger age groups.


Figure 1
View larger version (41K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1. Annual prevalence of antidepressant users among inhabitants in the County of Funen, Denmark, 1992–2004 (no. of users per 100 inhabitants per year). The estimated prevalence for the year 2004 is shown with 95% confidence intervals (CI). The number of subjects in each age group refers to 1 January 2004.

 
Exclusion of individuals 1 or 2 years prior to death decreased the prevalence estimates up to 2 percentage points, but time trends and the age relationship were not changed (data not shown).

In all age groups the use of antidepressants increased with proximity to death in the last 3 years of life (Table 1). This increase was most pronounced among males. Below the age of 95 years, about 33% of females and 25% of males had received antidepressants in the last 6 months of life. The same pattern of use was observed among people who died in earlier years (1995–98), but the levels were lower (data not shown).


View this table:
[in this window]
[in a new window]

  		Table 1. The proportion of people who received antidepressants during successive 6-month periods prior to date of death, among inhabitants dying in the period 2001–04

 
The change in the 1-year prevalence of SSRIs and TCAs in the different age groups is shown in Figure 2. In 1992, the TCAs dominated. For all age groups, however, the use of TCAs declined over time. The increase in the SSRIs has been most pronounced among the oldest age groups. In 2004, the 1-year prevalence was highest among older age groups whereas the opposite was true for the TCAs. The pattern shown in Figure 1 is mainly attributable to the increased use of SSRIs over the years.


Figure 2
View larger version (29K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 2. Annual prevalence of users of new (SSRIs) and old antidepressants (TCAs) among inhabitants in the County of Funen, Denmark, 1992–2004 (no. of users per 100 inhabitants per year).

 

   Discussion
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
key points
 Conflict of interest
 Notes
 References
 
This study provides an updated picture of the continuously increasing use of antidepressants among older people in routine clinical practice. The 1-year prevalence of antidepressants increases with age up to very high levels. As death approaches, the use of antidepressants increases regardless of age.

When evaluating the results, a number of issues need to be considered. Indications may differ with age, gender and calendar time and, furthermore, with physical health and comorbidity. Information on disease status or indication for the antidepressant treatment was unavailable as clinical outcome. It is a strength that the analysis is based on 13 years of prescriptions for a large unselected population. We used individual data from a pharmacy database covering all sales of antidepressants for all non-hospitalised persons in the county (considered representative of the Danish population) [13]. The validity and the completeness of the database are very high, and the population well defined by demographic data. Antidepressants purchased outside the county may result in underestimation of the prevalence estimates. Based on the percentage of antidepressants bought in the Funen County by people living in other areas (less than 5%) and the assumption that younger age groups are more mobile than older age groups we have, however, no reason to believe it invalidates our study. The present data allowed us to use the Kaplan–Meier method to estimate annual prevalences. Hence comparison across age groups was reliable despite high, age-related mortality rates. Time trend analyses may have been disturbed by change in official prescribing or redemption rules, but no significant changes occurred.

The positive age relationship observed is consistent with Canadian observations covering 1993–1997 [14], and cross-sectional data from Italy (2001) [15] and Finland (2002) [16]. These studies reported prevalence estimates based on the naïve approach and, thus, to some extent underestimated the prevalence in upper age groups. This study indicates that the use of antidepressants among older age groups has reached a level possibly higher than what may be expected from our knowledge of depression and treatment rates. The prevalence and age relationship of indications other than depression (anxiety and pain management for example) are uncertain. The indications may co-exist and are not always treatment requiring [17]. The positive age relationship observed was not attributable to the specific pattern of use at the end of life and has to be explained mainly by other mechanisms. Chronic disease, physical health problems and related disability may contribute to the age-related effects on prescribing rates [18, 19]. General practitioners have described late-life depression as part of a spectrum including loneliness, lack of social network, and reduction in functioning [20]. Antidepressants may be tried for relief of sadness and loneliness. Patients unlikely to benefit from the drugs are exposed to side-effects and the potential for interactions with other medications. In some cases, antidepressants may be prescribed instead of a more supportive intervention.

Generally little is known about depression during the last years of life other than it is associated with a host of prevalent medical conditions. Specific groups of patients have been studied, e.g. patients with congestive heart failure. Their depression scores have been shown to increase during the last months of life [21]. In palliative care settings the point prevalence of depression has been estimated at 15–30% [22]. Neither the patient nor the caregivers may be aware that death approaches, and remaining lifetime is relatively unpredictable for most older people. For terminally ill cancer patients clinicians' survival predictions have, however, been shown to be correlated with actual survival [23]. Our findings suggest that antidepressants are used for treatment of problems associated with end of life. Whether antidepressants are the appropriate treatment is still to be proved. The use of antidepressants in the last phase of life seemed independent of whether the patient dies at age 65 or 85 years which is consistent with a study showing that the number of days spent in hospital during the last years of life is independent of age [24]. What may explain the somewhat lower use of antidepressants among the 95+ year-olds during last years of life has to be investigated in further studies.

The 1-year prevalence of antidepressant prescribing differed substantially between people below and above 65 years, and specifically between the younger and the older of the 65+ year-olds. This study underlines the importance of focusing on older people in drug utilisation studies. Important knowledge may otherwise be lost [25]. Interpretation of prescribing patterns in inhomogeneous populations defined by wide age groups should be made cautiously. We furthermore conclude that by taking into account proximity to death one may gain new and clinically relevant insight into the use of specific drugs among older people.

The use of antidepressants for older people may be substantially different in healthcare systems other than the Danish. The consumption of antidepressants is, however, generally high in many countries and the levels in Denmark are not unique [14, 26]. This is the first population-based study aiming to analyse the use of antidepressants as death approaches, and the results deserve to be replicated in other contexts. The findings presented have important implications for society, clinicians and policy makers. The prescribing rates for older people are shown to be very high particularly close to death. Future studies may clarify the problems and diagnoses giving rise to the frequent prescribing for the old and especially as death approaches.


   key points
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 key points
Conflict of interest
 Notes
 References
 

  • Focus on older people has been lacking in most antidepressant drug utilisation studies and the use when death is approaching has never been analysed
  • The use of antidepressants continues to increase in all age groups
  • Among the 65+ year-olds, the use of antidepressants increases with age to levels much higher than what may be expected
  • The use of antidepressants increases during the last 3 years of life irrespective of age: 33% of females and 25% of males had antidepressants during their last 6 months of life
  • It is important to explore the reasons for prescribing antidepressants for so many older people, and especially the problems giving rise to treatment as death approaches.


   Conflict of interest
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 key points
 Conflict of interest
Notes
 References
 
None.


   Acknowledgements
 
We thank Secretary Lise Stark for proofreading the manuscript. DGH study was supported by grant from The Novo Foundation.


   Notes
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 key points
 Conflict of interest
 Notes
References
 
1 Anatomical classification by WHO (10) giving a code for each specific pharmacological substance. Back


   References
Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 key points
 Conflict of interest
 Notes
 References
 

  1. Henriksson S, Boethius G, Hakansson J, Isacsson G. Indications for and outcome of antidepressant medication in a general population: a prescription database and medical record study, in Jamtland county, Sweden, 1995. Acta Psychiatr Scand (2003) 108:427–31.[CrossRef][Web of Science][Medline]
  2. Rosholm JU, Nybo H, Andersen RK, et al. Hyponatraemia in very old nonhospitalised people: association with drug use. Drugs Aging (2002) 19:685–93.[CrossRef][Web of Science][Medline]
  3. de Abajo FJ, Rodriguez LA, Montero D. Association between selective serotonin reuptake inhibitors and upper gastrointestinal bleeding: population based case-control study. BMJ (1999) 319:1106–9.[Abstract/Free Full Text]
  4. Liu B, Anderson G, Mittmann N, To T, Axcell T, Shear N. Use of selective serotonin-reuptake inhibitors of tricyclic antidepressants and risk of hip fractures in elderly people. Lancet (1998) 351:1303–7.[CrossRef][Web of Science][Medline]
  5. Stek ML, Gussekloo J, Beekman AT, Van Tilburg W, Westendorp RG. Prevalence, correlates and recognition of depression in the oldest old: the Leiden 85-plus study. J Affect Disord (2004) 78:193–200.[CrossRef][Web of Science][Medline]
  6. Hirschfeld RM, Keller MB, Panico S, et al. The National Depressive and Manic-Depressive Association consensus statement on the undertreatment of depression. JAMA (1997) 277:333–40.[Abstract/Free Full Text]
  7. Bergdahl E, Gustavsson JM, Kallin K, et al. Depression among the oldest old: the Umea 85+ study. Int Psychogeriatr (2005) 17:557–75.[CrossRef][Web of Science][Medline]
  8. Rosholm JU, Hallas J, Gram LF. Outpatient utilization of antidepressants: a prescription database analysis. J Affect Disord (1993) 27:21–8.[CrossRef][Web of Science][Medline]
  9. Hansen DG, Søndergaard J, Vach W, Gram LF, Rosholm JU, Kragstrup J. Antidepressant drug use in general practice: Inter-practice variation and association with practice characteristics. Eur J Clin Pharmacol (2003) 59:143–9.[Web of Science][Medline]
  10. Middleton N, Gunnell D, Whitley E, Dorling D, Frankel S. Secular trends in antidepressant prescribing in the UK, 1975–1998. J Public Health Med (2001) 23:262–7.[Abstract/Free Full Text]
  11. Barbui C, Broglio E, Laia AC, et al. Cross-sectional database analysis of antidepressant prescribing in Italy. J Clin Psychopharmacol (2003) 23:31–4.[CrossRef][Web of Science][Medline]
  12. Glynn RJ, Knight EL, Levin R, Avorn J. Paradoxical relations of drug treatment with mortality in older persons. Epidemiology (2001) 12:682–9.[CrossRef][Web of Science][Medline]
  13. Gaist D, Sorensen HT, Hallas J. The Danish prescription registries. Dan Med Bull (1997) 44:445–8.[Web of Science][Medline]
  14. Mamdani M, Herrmann N, Austin P. Prevalence of antidepressant use among older people: population-based observations. J Am Geriatr Soc (1999) 47:1350–3.[Web of Science][Medline]
  15. Percudani M, Barbui C, Fortino I, Petrovich L. Antidepressant drug prescribing among elderly subjects: a population-based study. Int J Geriatr Psychiatry (2005) 20:113–8.[CrossRef][Web of Science][Medline]
  16. Hartikainen S, Klaukka T. Use of psychotropics is high among very old people. Eur J Clin Pharmacol (2004) 59:849–50.[CrossRef][Web of Science][Medline]
  17. Flint AJ. Generalised anxiety disorder in elderly patients: epidemiology, diagnosis and treatment options. Drugs Aging (2005) 22:101–14.[CrossRef][Web of Science][Medline]
  18. Roberts RE, Kaplan GA, Shema SJ, Strawbridge WJ. Does growing old increase the risk for depression? Am J Psychiatry (1997) 154:1384–90.[Abstract]
  19. Blazer DG, Hybels CF, Fillenbaum GG, Pieper CF. Predictors of antidepressant use among older adults: have they changed over time? Am J Psychiatry (2005) 162:705–10.[Abstract/Free Full Text]
  20. Burroughs H, Lovell K, Morley M, Baldwin R, Burns A, Chew-Graham C. Justifiable depression: how primary care professionals and patients view late-life depression? A qualitative study. Fam Pract (2006) 23:369–77.[Abstract/Free Full Text]
  21. Levenson JW, McCarthy EP, Lynn J, Davis RB, Phillips RS. The last six months of life for patients with congestive heart failure. J Am Geriatr Soc (2000) 48:S101–9.[Medline]
  22. Hotopf M, Chidgey J, Addington-Hall J, Ly KL. Depression in advanced disease: a systematic review Part 1. Prevalence and case finding. Palliat Med (2002) 16:81–97.[Abstract/Free Full Text]
  23. Glare P, Virik K, Jones M, et al. A systematic review of physicians' survival predictions in terminally ill cancer patients. BMJ (2003) 327:195.[Abstract/Free Full Text]
  24. Dixon T, Shaw M, Frankel S, Ebrahim S. Hospital admissions, age, and death: retrospective cohort study. BMJ (2004) 328:1288.[Abstract/Free Full Text]
  25. McMurdo ME, Witham MD, Gillespie ND. Including older people in clinical research. BMJ (2005) 331:1036–7.[Free Full Text]
  26. Medicines Consumption in the Nordic Countries 1999–2003 (2004) Copenhagen: NOMESCO Nordic Medico Statistical Committee. 72.
Received 21 November 2006; accepted in revised form 18 April 2007.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
36/4/449    most recent
afm056v1
Right arrow E-letters: Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when E-letters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Hansen, D. G.
Right arrow Articles by Vach, W.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hansen, D. G.
Right arrow Articles by Vach, W.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?