Consideration of Personal and Environmental Variables in the Measurement of Psychophysiological Stress Response to Different Environmental Settings

Purpose: To test the design of a pilot study measuring urban environmental effects on psychological and physiological measures of stress.

Methods: A four-way crossover design was used. Fifteen participants passively experienced each of 4 outdoor urban settings (i.e. Very Natural (VN), Mostly Natural (MN), Mostly Built (MB), and Very Built (VB)) on separate days in randomized order for 20 minutes each. Pre- and post-exposure data were collected via saliva and self-report questionnaires. Data regarding participants' pre-existing stress state, environmental self-identity, and perceived restorativeness and aesthetic appreciation of each setting were collected as co-variates. Environmental variables (i.e. ambient light, sound, temperature, humidity) were also included in the study design, and qualitative post-exposure data were collected to inform future study design.

Results: Compliance and attendance were very high for this repeat-measures study (59 of 60 visits). ANCOVA revealed near-significant between-setting differences for changes in subjective stress (F_3,40.84=2.670, p=0.060), tension-anxiety (T-A: F_3,41.21=2.986, p=.043) and salivary alpha-amylase (sAA: F_3,38.3=1.69, p=0.186), but not for salivary cortisol. Post-hoc analyses revealed significant differences, primarily between the VN and MB settings (stress: p=0.008; T-A: p=0.007; sAA: p=0.033). Co-variate measures included in analyses were shown to have important moderating effects. Equipment error prevented the inclusion of environmental data the analyses. However, qualitative data demonstrated that noise, as well as presence of non-study personnel, familiarity with the setting, and presence of odors, influenced participants' subjective experience of each setting.

Conclusion: Measurement of environmental stressors in a field setting is challenging but possible. The need to consider multiple environmental and personal variables necessitates the use of complex, multivariate analysis techniques. Special attention toward the natural/built content of urban environments may be fruitful in understanding how outdoor spaces can be utilized to beneficially modulate the stress response.

Contact: Kurt Beil, kbeil@ncnm.edu