Chronic and remitting trajectories of depressive symptoms in the elderly. Characterisation and risk factors

Isabelle Carrière, A. Farré, Cécile Proust-Lima, J. Ryan, M. L. Ancelin, K. Ritchie

Research output: Contribution to journalArticleResearchpeer-review

13 Citations (Scopus)


Background: In elderly general population sub-syndromal clinically significant levels of depressive symptoms are highly prevalent and associated with high co-morbidity and increased mortality risk. However changes in depressive symptoms over time and etiologic factors have been difficult to characterise notably due to methodological shortcomings. Our objective was to differentiate trajectories of depressive symptoms over 10 years in community-dwelling elderly men and women using statistical modelling methods which take into account intra-subject correlation and individual differences as well as to examine current and life-time risk factors associated with different trajectories. Methods: Participants aged 65 and over were administered standardised questionnaires and underwent clinical examinations at baseline and after 2, 4, 7 and 10 years. Trajectories over time of the Center for Epidemiologic Studies Depression scores were modelled in 517 men and 736 women separately with latent class mixed models which include both a linear mixed model to describe latent classes of trajectories and a multinomial logistic model to characterise the latent trajectories according to baseline covariates (socio-demographic, lifestyle, clinical, genetic characteristics and stressful life events). Results: In both genders two different profiles of symptom changes were observed over the 10-year follow-up. For 9.1% of men and 25% of women a high depressive symptom trajectory was found with a trend toward worsening in men. The majority of the remaining men and women showed decreasing symptomatology over time, falling from clinically significant to very low levels of depressive symptoms. In large multivariate class membership models, mobility limitations [odds ratio (OR) = 4.5, 95% confidence interval (CI) 1.6–12.9 and OR = 4.9, 95% CI 2.3–10.7, in men and women respectively], ischemic pathologies (OR = 2.9, 95% CI 1.0–8.3 and OR = 3.1, 95% CI 1.0–9.9), and recent stressful events (OR = 4.5, 95% CI 1.1–18.5, OR = 3.2, 95% CI 1.6–6.2) were associated with a poor symptom course in both gender as well as diabetes in men (OR = 3.5, 95% CI 1.1–10.9) and childhood traumatic experiences in women (OR = 3.1, 95% CI 1.6–5.8). Conclusions: This prospective study was able to differentiate patterns of chronic and remitting depressive symptoms in elderly people with distinct symptom courses and risk factors for men and women. These findings may inform prevention programmes designed to reduce the chronic course of depressive symptomatology.

Original languageEnglish
Pages (from-to)146-156
Number of pages11
JournalEpidemiology and Psychiatric Sciences
Issue number2
Publication statusPublished - Apr 2017
Externally publishedYes


  • Aged
  • cohort studies
  • depressive symptoms
  • risk factors

Cite this