Three versions of Perceived Stress Scale: validation in a sample of Chinese cardiac patients who smoke

Abstract

Abstract Background Smoking causes heart disease, the major cause of death in China and Hong Kong. Stress is one major trigger of smoking and relapse, and understanding stress among smoking cardiac patients can therefore help in designing effective interventions to motivate them to quit. The objective of this study was to examine the psychometric properties of the Perceived Stress Scale (PSS), and to compare the appropriateness of the three versions of the scale (PSS-14, PSS-10, and PSS-4) among Chinese cardiac patients who were also smokers. Methods From March 2002 to December 2004, 1860 cardiac patients who smoked were recruited at the cardiac outpatient clinics of ten acute hospitals in Hong Kong, and 1800 questionnaires were analysed. Participants completed a questionnaire including the PSS, nicotine dependence and certain demographic variables. The psychometric properties of the PSS were investigated: construct validity using confirmatory factor analysis, reliability using Cronbach's alpha and concurrent validity by examining the relationship with smoking- and health-related variables. Results For all the three versions of the PSS, confirmatory factor analyses corroborated the 2-factor structure of the scale, with the positive and negative factors correlating significantly and negatively to a moderate extent ( r < -0.5), and high Cronbach's alpha values for the two subscales (alpha > 0.5). All the correlations of the two subscales and the smoking- and health-related variables were statistically significant and in the expected directions although of small magnitudes, except daily cigarette consumption. Conclusions The findings confirmed the satisfactory psychometric properties of all three Chinese versions of PSS. We recommend the use of PSS-10 for research which focuses on the two components of perceived stress, as it shows a higher reliability; and the use of PSS-4 if such partition is not essential and space for multiple measures is limited.