Sedative load among community-dwelling people aged 75 years and older: a population-based study.

Background: Drugs with sedative properties are among the most widely used drugs in community-dwelling older people. Use of sedative drugs has been associated with falls and fractures, cognitive and memory impairment and impaired physical function among older people. The sedative load model has been developed to quantify the cumulative effect of taking multiple drugs with sedative properties.

Objective: The objective of the study was to investigate factors associated with sedative load among community-dwelling older people, using data collected as part of the Finnish Geriatric Multidisciplinary Strategy for the Good Care of the Elderly (GeMS) study.

Methods: The GeMS study was a randomized, comparative study that evaluated a model for geriatric assessment, care and rehabilitation using a study sample of 1000 persons aged≥75 years who were living in Kuopio, Finland. Of these, 700 people consented to participate and were community-dwelling. Demographic, diagnostic and drug use data (both regular and when-required drugs) were elicited during nurse interviews. For the current analysis, sedative load was computed using a previously published model, in which drugs taken on a regular and when-required basis were classified into one of four groups according to their sedative potential. Group 1 included primary sedatives (sedative rating 2) and group 2 included drugs with sedation as a prominent side effect (sedative rating 1). Each participant's sedative load was calculated by summing the sedative ratings of group 1 and 2 drugs. Logistic regression models were used to investigate factors associated with sedative load.

Results: Twenty-nine percent of participants (n=205) had a sedative load of ≥1 (i.e. used one or more drugs with sedative properties), and 22% (n=158) had a sedative load of ≥2 (i.e. used either one primary sedative or two drugs with sedation as a prominent adverse effect or preparations with a sedating component) when considering regularly used drugs. A sedative load of ≥2 that related to regularly used drugs was associated with female sex (odds ratio [OR] 1.65; 95% CI 1.02, 2.67), poor self-perceived health (OR 2.06; 95% CI 1.25, 3.38), impaired instrumental activities of daily living [IADL] (OR 1.89; 95% CI 1.18, 3.01) and often feeling lonely (OR 4.72; 95% CI 2.15, 10.40). The same factors remained significantly associated with a sedative load of ≥2 after drugs used on a when-required basis were included in the analyses.

Conclusions: The advantages of the sedative load model were that it included drugs with sedative properties prescribed for somatic diseases, described cumulative exposure to drugs that exert sedative effects through multiple mechanisms in the CNS, and incorporated a sedative rating for each drug. In an older population, female sex, impaired IADL, poor self-perceived health, and loneliness were associated with higher sedative load. Clinicians should remain cognizant of these factors when reviewing drug regimens and targeting interventions to optimize sedative use.