Abstract
INTRODUCTION: Despite the availability of validated psychometrics tools to assess depression, there has not been any validated and reliable tool established to test perceived stress among Sri Lankans. The objective of this study is to test the validity and reliability of the Sinhalese Version of the Sheldon Cohen Perceived Stress Scale. MATERIALS AND METHODS: Standard and systematic procedures were adopted to translate the original English version of the Perceived Stress Scale-10 questionnaire into Sinhalese. Consecutive sampling was employed to recruit the Type 2 Diabetes mellitus (T2DM) sample (n = 321), and a convenient sampling was used to recruit the Age and Sex matched Healthy Controls (ASMHC) (n = 101) and the Healthy Community Controls (HCC) groups (n = 75). Cronbach alpha was used to assess internal consistency and reliability was determined using test-retest method utilizing Spearman's correlation coefficient. Sensitivity was evaluated by comparing the mean scores of the Sinhalese Perceived Stress Scale (S-PSS-10) and Sinhalese Patient Health Questionnaire (S-PHQ-9) scores. Post-hoc comparisons were done using Bonferroni's method. Mean scores were compared between the T2DM, ASMHC, and HCC groups using the independent t-test. Explanatory Factor Analysis (EFA) was conducted using the principal component and Varimax rotation while the Confirmatory Factor Analysis (CFA) was performed to assess the goodness-of-fit of the factor structure extracted from the EFA. Concurrent validity was assessed using the Pearson correlation between the S-PSS-10 and Patient Health Questionnaire measured by S-PHQ-9 (p < 0.05). RESULTS: Cronbach alpha values of the three groups T2DM, ASMHC and HCC were 0.85, 0.81, and 0.79, respectively. Results of the ANOVA test suggested that there was a significant difference in the mean scores between groups (p < 0.00). EFA analysis revealed the existence of two factors with eigenvalues greater than 1.0. The factor loadings for the items ranged from 0.71-0.83. The CFA analysis demonstrated a good model fit for the two-factor model S-PSS-10. The S-PSS-10 significantly correlated with S-PHQ-9, indicating an acceptable concurrent validity. CONCLUSION: Findings revealed that the S-PSS-10 questionnaire can be used to screen perceived stress among the majority of the Sri Lankan Sinhalese-speaking population specially with chronic illnesses. Further studies with higher sample sizes across different populations would enhance the validity and reliability of S-PSS-10.