Prevalence, patterns and predictors of depression treatment among community-dwelling older adults with stroke in the United States: A cross sectional study

Sandipan Bhattacharjee, Majed Al Yami, Sawsan Kurdi, David Rhys Axon

Research output: Contribution to journalArticle

1 Scopus citations


Background: Depression is one of the most common psychiatric conditions among stroke survivors and is associated with several negative health outcomes. However, little is known about the depression treatment patterns among stroke survivors. The objective of this study was to examine national-level prevalence, patterns and predictors of depression treatment among community-dwelling stroke survivors. Methods: This study adopted a retrospective, cross-sectional study design using multiple years of Medical Expenditure Panel Survey (MEPS) (2002-2012) data. The study population consisted of older adults (age ≥ 50 years) who (i) were stroke survivors (ICD-9-CM codes of 430-438), (ii) did not die during the calendar year, and (iii) had co-occurring depression (ICD-9-CM code of 296.xx, or 311.xx). Depression treatment, identified by antidepressant medication and/or psychotherapy use, was the dependent variable of this study. Multinomial logistic regression analysis was conducted to examine the association of individual level factors with depression treatment among stroke survivors with co-occurring depression. Results: The final study sample consisted 370 (unweighted) community-dwelling older adults with self-reported stroke and depression. The prevalence of co-occurring depression among stroke survivors was 22.03% [95% Confidence Interval (CI) 19.7-24.4%]. An overwhelming majority (87.6%) of stroke survivors with co-occurring depression reported some form of depression treatment. Antidepressants only and combination therapy was reported by 74.8% (95% CI, 71.6-78.0%] and 12.8% (95% CI, 10.5-15.1%) by stroke survivors with co-occurring depression respectively. Approximately, 61% of stroke survivors with co-occurring depression reported using SSRIs, followed by SNRIs (15.2%), miscellaneous antidepressants (12.1%), TCAs (9.8%), phenylpiperazine antidepressants (5.2%), and tetracyclic antidepressants (4%). Sertraline (15.8, 95% CI, 12.7-19.0%) had the highest reported use among individual antidepressants. Conclusions: Vast majority (nearly 90%) of the study sample received some form of depression treatment and several individual level factors (such as age, education) were associated with the report of depression treatment use. Future longitudinal studies are warranted to assess the comparative treatment benefits of antidepressants, psychotherapy and their combination. Healthcare providers should carefully assess the risks and benefits of antidepressant (such as SSRIs or TCAs) use in this vulnerable population prior to their use.

Original languageEnglish (US)
Article number130
JournalBMC Psychiatry
Issue number1
Publication statusPublished - May 16 2018



  • Antidepressants
  • Depression
  • Psychotherapy
  • Stroke

ASJC Scopus subject areas

  • Psychiatry and Mental health

Cite this