The article you requested is

The Relationships of Environment and Personal Characteristics to Agitated Behaviors in Nursing Home Residents With Dementia

J Clin Psychiatry 2012;73(3):392-399

Objective: To explore the impact of personal attributes, environmental attributes, and the presentation of 9 categories of stimuli on agitation in nursing home residents with dementia.

Method: Participants in this randomized, controlled, observational cross-sectional study were 193 residents of 7 nursing homes, all with a diagnosis of dementia, for whom we obtained data pertaining to cognitive functioning (via the Mini-Mental State Examination), performance of activities of daily living (Minimum Data Set), and role-identity/activities of past interest (Self-Identity Questionnaire). Environmental attributes (eg, noise, lighting) and direct observations of agitation (primary outcome) were recorded via the Agitation Behavior Mapping Inventory. Engagement was measured using the Observational Measurement of Engagement. Both agitation and engagement were assessed for each stimulus (including a control condition).

Results: Univariate findings (ie, for 1 explanatory variable at a time) showed agitation to be related to several personal attributes—ie, female gender was related to verbal agitation (P < .0001); low cognitive function was related to total, verbal, and physical agitation (P < .001 for each); low performance of activities of daily living was significantly related to all types of agitation (P < .01 for total agitation and P < .05 for each type of agitation); and unclear speech was significantly related to total agitation (P < .01). Eight of the 9 stimulus categories were significantly related to decreased levels of agitation, with ORs ranging from 0.37 (live human stimuli, P < .001) to 0.79 (inanimate social stimuli, P < .05). Higher levels of engagement were related to lower levels of agitation (P < .001 for total agitation). In the multivariate analyses, higher cognitive function (P < .001), male gender (P < .05), level of engagement with stimuli (eg, duration of engagement for 3 minutes or longer, P < .05), and all 9 stimulus categories, with the exception of music, were independently predictive of lower levels of agitation (P < .001).

Conclusions: The finding that both type of stimuli and engagement level with the stimuli were significant predictors of agitation underscores the importance of engagement as a determinant of agitation levels.

J Clin Psychiatry