Drugs are one of the presumed causes of falls in the elderly. Research has not addressed the pharmacologic mechanisms whereby drugs cause falls. Most medicines associated with falls have anticholinergic activity, but other possible mechanisms are sedation and postural hypotension. This is a pilot study t ascertain plausibility of the hypothesis that anticholinergic activity is a pharmacologic mechanism of drug-induced falls. This study also examines the relationship of postural sway (a measure of static balance), dynamic balance, and fear of falling with pharmacologic properties of drugs.
The aims of this project are to describe falls associated with medications; estimate the fall risk associated with anticholinergic drugs; ascertain the amount of variance in the dependent variables (postural sway, dynamic balance, and fear of falling) explained by selected predictor variables (anticholinergic dose, sedation, and postural sway); and compare postural sway at peak and trough of anticholinergic activity. The study is a longitudinal descriptive correlational design. After a preliminary study of 10 subjects to refine study procedures, 110 elderly taking drugs associated with fall with be recruited from community locations. Subjects will be assessed on the predictor variables of anticholinergic dosage (in atropine equivalents computed across all drugs), postural hypotension, and sedation (measured by the Mood Rating Scale and the Digit Symbol Substitution Test), as well as on the dependent variables of postural sway (area of the ellipse, sway velocity, and lateral sway measured using biochemical force platform), functional dynamic balance (Berg Balance Scale) and fear of falling (Modified Falls Efficacy Scale). Fall events (falls, near falls) during 12 months and time to first fall event will be ascertained by fall diaries, postcards, and telephone interview. A subsample of 40 patients taking either drugs with anticholinergic properties or taking no drugs with anticholinergic properties will be compared on sway at projected time of peak and trough drug levels. Analysis will include descriptive statistics, logistic regression, content analysis, stepwise multiple regression, and repeated measures ANOVA. Explanation of significance variance in falls or postural sway by anticholinergic dose and increased postural sway at time estimated peak drug levels would indicate anticholinergic activity is a tenable mechanism of drug-induced falling in the elderly. If there is anticholinergic dose effect, elderly adults with high fall risk should be prescribed alternative medications with similar therapeutic effects, but smaller impact on falls.
