Document Type


Date of Degree

Spring 2013

Degree Name

PhD (Doctor of Philosophy)

Degree In


First Advisor

Specht, Janet K

First Committee Member

Kolanowski, Ann

Second Committee Member

Mobily, Paula

Third Committee Member

Kleiber, Charmaine

Fourth Committee Member

Herr, Keela A


Alzheimer's disease is a form of dementia associated with disturbing and disruptive behaviors that account for many negative health and well-being outcomes, including declines in functional status, social engagement, and physical activity (Lyketsos, 2007). These behavioral consequences diminish patients' quality of life (QoL) and increase caregiver burden and the cost of care, often ultimately necessitating that patients be placed in a nursing home (Murman and Colenda, 2005).

Nature can profoundly affect people's health, well-being, and QoL; indeed, it is an old concept that the healing properties of nature can be used therapeutically. Accordingly, naturally restorative environmental (NRE) interventions stimulate one or more of the senses using natural things: elements of the earth that are living and animate, geographic, or solar and climatic (Gibson, Chalfont, Clarke, Torrington, and Sixsmith, 2007). For persons with dementia, interventions that incorporate NRE elements have demonstrated a variety of benefits, including decreased agitation; less use of psychotropic drugs); normalization of the circadian rhythm; and enhanced sociability, affect, cognitive capacity, and attention (Detweiler, Murphy, Kim, Meyers, and Ashai, 2009; Colenda, Cohen, McCall, and Rosenquist, 1997; LeGrace, 2002). Self-reported improvements in well-being, quality of life, and participation in meaningful activities have also been documented (Collins and O'Callaghan, 2008; Duggan, Blackman, Martyr, and Van Schaik, 2008; Nowak and Davis, 2011). Thus, NREs provide caregivers potential options for addressing physical, spiritual, psychological, and social needs, while at the same time, affecting behavioral responses. These widespread benefits justify further investigation and clarification.

Despite the rich potential of NRE interventions for treating dementia, the research to support NRE use has not been synthesized and defined in terms of specific behaviors that may be affected, their dosage, the optimal NRE settings, and other specific characteristics. Further research is needed to develop the most effective interventions. The purpose of this dissertation is to produce a comprehensive meta-analysis of the studies, both published and unpublished, that detail the use of NRE in interventions for behaviors and QoL in dementia care. A meta-regression was conducted of moderator variables to guide development of NRE interventions for dementia care. Additionally, the characteristics of different types of programs were synthesized. Data from thirty three articles were pooled for effect size (ES) estimates on two outcomes: disruptive behaviors and quality of life. In two-group comparisons, treatment and control, an ES = 0.484 + .138, k= 17, CI (0.215, 0.745) favored the interpretation that disruptive behaviors were attenuated with NRE interventions. Findings were higher in single, pre- post-test design studies, with an ES= 0.758+ 0.109. k= 7, CI (0.544, 0.973) for diminishing disruptive behaviors. In the two-group comparisons assessing the quality of life outcome, the ES= 0.579 + 0.171, k= 10, CI (0.243, 0.915); for single group designs, the ES= 1.347+ 0.256, k=7, CI (0.020, 0.719). Thus both design analyses indicated improved measures of QoL for persons with dementia. Moderator analysis by type of NRE, two-group design, showed statistically significant lessening in disruptive behaviors using aromatherapy, but not bright light or horticulture therapy. Better QoL was shown independently by the moderators, aromatherapy and horticulture therapy, but not bright light therapy. When single group analysis was done for each type of NRE, they all independently showed significance for both behaviors and QoL; except there were no single group designs in AT or BLT. Despite the considerable heterogeneity of the interventions, individual moderators all showed potential benefits, in a variety of settings, and in different contexts.


Dementia, meta-analysis, Naturally restorative environment, nonpharmacological interventions


ix, 149 pages


Includes bibliographical references (pages 126-149).


Copyright © 2013 Ann L. Gibbs Bossen

Included in

Nursing Commons